WO2016193553A1

WO2016193553A1 - System for aggregating video content, comprising a module for measuring and managing user votes

Info

Publication number: WO2016193553A1
Application number: PCT/FR2015/000110
Authority: WO
Inventors: Anthony ORNECQ
Original assignee: Ornecq Anthony
Priority date: 2015-06-03
Filing date: 2015-06-03
Publication date: 2016-12-08

Abstract

The invention relates to a system (1) for aggregating video content, comprising: a remote server (10) having video sequences (α); user terminals (11) provided with human-machine interfaces (110) and connected to said server (10) and to an aggregator (12) of video content including a reading module (121) suitable for reading (L1) the video sequences (α). According to the invention, the aggregator (12) includes a voting module (122) designed to: associate (V1) each current video sequence (α_i) with a voting session (S_i); count (V2) the votes (v) cast (H4) by the human-machine interfaces (110); obtain (V3) a current result (β₁) representative of the votes (v) counted at a selective time (τ₁); compare (V4) said current result (β₁) with a threshold (σ₁); and supply (V5) a warning signal (φ₁) to the reading module (121) when the current result (β₁) exceeds said threshold (σ₁).

Description

VIDEO CONTENT AGGREGATION SYSTEM WITH MEASUREMENT MODULE AND MANAGEMENT OF USER'S VOTES

The present invention generally relates to the field of on-demand video broadcasting of social network users and, more particularly, to the field of selectively managing said video streaming on demand through a video content aggregator.

More precisely, the invention relates, according to a first aspect, to a video content aggregation system comprising:

A remote server comprising a database with video content consisting of video sequences, and user terminals connected to each other and to the remote server via a telecommunication network via a network aggregator. video content.

In this aggregation system, the user terminals each comprise a human-machine interface adapted to issue orders to the aggregator and to display the video sequences read by the aggregator. As for the aggregator, it includes a translation module adapted to translate the commands transmitted by the human-machine interfaces of user terminals into video content requests to the remote server. In turn, the remote server is adapted to generate a stream of video footage to the aggregator in response to its video content requests. Finally, the aggregator further comprises a video content reading module adapted to read, one after another, the video sequences from the stream of video sequences.

Users of a telecommunication network, for example, a global international telecommunication network called "internet", and in particular those among them who are members of one of the social networks, can organize themselves in a community of viewers-amateurs of television on demand, and enjoy the aggregation system known and described above. So, these spectators-amateurs of the television on demand can only watch the video sequences selectively programmed by the community members, that is, by viewers themselves, with the remote server with the video content base, using the corresponding orders issued by the user terminals. However, the expected operation of this known aggregation system is based on a "serial" reading mode, that is, one after the other, video sequences selected by the members of the community. However, this mode of serial playback of video sequences may have deficiencies due, for example, to a need, for each viewer (given member of the community), to view video sequences without interest programmed by other viewers (who are members of the same community), before reaching a sequence relevant to said viewer and, ultimately, to the community of all viewers as a whole. Such a deficiency results in a waste of time for each given member of the community. Thus, despite the fact that this video broadcast based on the known aggregation system is intrinsically selective (because it is necessarily the viewers themselves who choose the video sequences subsequently broadcast via the human-machine interfaces of the user terminals ), it leads to the unsatisfactory result in terms of waiting time required (for each spectator of the community) before reading the relevant sequence, this unsatisfactory result being ultimately similar in these terms to that obtained by a conventional television a priori non-selective (because the video sequences are presented according to a program made by third parties that are not necessarily part of the community of spectators). This poses a problem in controlling the selective management of video on demand broadcasting.

The present invention, which is based on this original observation, is primarily intended to provide a video content aggregation system to at least reduce a previously mentioned limitation.

For this purpose, the video content aggregation system, which is also consistent with the generic definition given in the preamble above, is essentially characterized in that the aggregator further comprises a voting module linked with the reading module and designed to:

Associating each video sequence currently being read by the reading module with a current voting session specific to said video sequence

5, human-machine interfaces of user terminals being adapted to:

o display the current voting session together with the corresponding current video sequence,

o issue votes to the aggregator during the current voting session,

• continuously count the votes sent by the human-machine interfaces to the aggregator, this continuous counting of votes taking place in real time without interruption, by adding them to each other, as and when they arrive of votes in the module

1 5 of voting, according to a first predetermined rule previously stored by the voting module,

Obtain a first current result representative of the votes counted at a first selective moment of the current voting session, the selection of the first selective moment taking place according to a second rule

20 preset previously stored by the voting module,

Comparing this first current result of the current voting session with a first predetermined threshold previously stored by the voting module,

Provide a first alert signal to the read module when the first current result exceeds the first predetermined threshold.

Thanks to this voting module, the video content aggregation system according to the invention (and, in particular, the reading module of the aggregation system), can, via the first configurable warning signal, have a first return information regarding an objectively reserved reception by viewers to a sequence being played visible by viewers on their respective terminals. Thus, the relevance of this current sequence for a whole community of viewers can be measured in real time. This makes it possible to manage the reading module more finely by the video content aggregation system according to the invention and ultimately contributes to a more effective control of the video broadcast on demand.

According to a second aspect, the invention relates to a method for aggregating video content implemented by the video content aggregation system according to the invention.

Other characteristics and advantages of the invention will emerge clearly from the description which is given hereinafter, by way of indication and in no way limitative, with reference to the appended drawings, in which:

FIG. 1 represents a (functional) diagram of a variant of a video content aggregation system according to the invention.

As previously announced and illustrated in FIG. 1, the invention relates to a video content aggregation system and a method implemented by this aggregation system.

In an alternative embodiment illustrated in FIG. 1, the invention relates to a video content aggregation system 1 comprising:

A remote server 10 comprising a database 100 with a video content consisting of video sequences,

• User terminals 11 connected to each other and to the remote server 10 via a telecommunications network 2 via an aggregator 12 of video content.

The user terminals 1 1 each comprise a man-machine interface 1 10 adapted to transmit H1 orders p to the aggregator 12 and display H2 the video sequences read by the aggregator 12. The aggregator 2 comprises a 120 translation module adapted to translate T1 the orders p emitted by the human-machine interfaces 1 10 of user terminals 1 1 in P requests for video content to the remote server 10. The remote server 10 is adapted to generate F1 a stream of video sequences a to the aggregator 12 in response to its 5 requests P of video content. The aggregator 12 further comprises a reading module 121 of video content adapted to read L1, one after the other, the video sequences a from the stream of video sequences.

The remote servers 10 mentioned above may belong to community web platforms (video content providers) known for sharing and viewing video sequences made by user-viewers. Their respective databases 100 with the video contents consisting of video sequences are available, for example, via the following links: www.youtube.com; www.vimeo.com, www.dailymotion.com; www.vouku.com; www.tudou.com.

According to the invention, the aggregator 12 further comprises a voting module 122 linked with the reading module 121 and designed to:

Associating each video sequence V1 being read by the reading module 121 with a current voting session S specific to said current video sequence, the human-machine interfaces 1 10 of 0 user terminals 1 being adapted to:

o display H3 the voting session S, running together with the video sequence a, corresponding current, o issue H4 votes v to the aggregator 12 during the voting session S, current, 5 · count V2 continuously votes v issued H4 by the human-machine interfaces 1 10 to the aggregator 12, this counting of continuous votes v operates in real time without interruption, by adding them to each other, as and when v vote arrivals in the voting module 122, according to a predetermined first rule Ri previously stored by the voting module 122,

• obtain V3 a first result βι current representative votes v counted at a first time τι selective vote session If current, the choice of the first moment τι selective operating according to a second predetermined rule R2 previously stored by the voting module 122

Compare V4 this first result βι current of the voting session S, current with a first threshold ai predetermined previously stored by the voting module 122,

• supply V5 a first warning signal φι to the reading module 121 when the first result βι current exceeds the first threshold σι predetermined.

Users of the telecommunication network 2, for example, of a global international telecommunication network called "internet", and in particular those among them who are members of one of the social networks and have user terminals 1 1 , can organize themselves in community of spectators-amateurs of television on demand, and take advantage of the aggregation system 1 according to the invention and described above. Thus, these amateur-on-demand television viewers can only watch video sequences ordered (and, therefore, selectively programmed) by community members, that is, by viewers. same, with the remote server 10 provided with the base 100 of video content, with the corresponding p commands issued by the user terminals 1 1 via the human-machine interfaces 1 10.

Each human-machine interface 1 10 may comprise a communication means (not shown in FIG. 1). This communication means can be adapted to any type of incoming and outgoing signals, for example, to visual communication, and / or sound, and / or tactile, and / or odorous. In a preferred embodiment of the invention, the user terminals 1 1 can be presented as computers, including laptops, tablets, smart phones and other equipment provided with a computer, a screen (including a touch screen), a keyboard (including a specific keyboard, for example, a Braille keyboard), and / or a microphone, and / or a generator 2. Thus, each on-demand viewer can communicate with the aggregation system 1 using a user terminal 1 which is assigned to him or her. clean.

Preferably, the voting module 122 is designed to:

To continue V6 the continuous counting of the votes v during a predetermined time interval calculé calculated from the first n-selective moment, this predetermined time interval étant being previously memorized by the voting module 122,

15 · obtain V7 a second result β ₂ current representative of the votes v counted at the end of the time interval Θ predetermined of the voting session S, current, the votes v counted during the time interval Θ predetermined adding to those obtained at the first n selective moment, 0 · compare V8 this second result β ₂ current with a second threshold σ ₂ predetermined different from the first threshold σι predetermined,

• supply V9 a second warning signal φ ₂ to the reading module 121 when the second result β ₂ current exceeds the second threshold σ ₂ predetermined.

Thanks to this arrangement, the aggregation system 1 according to the invention

(and, in particular, the reading module 121 of the aggregation system 1), can, via the second configurable warning signal φ ₂ , have a second feedback information concerning an objectively reserved reception by the spectators to a video sequence a, being read by the module of 121 and viewable by viewers on their respective user terminals 1 1. This second feedback information is added to the first feedback information (obtained, as explained above, thanks to the first warning signal φι) which makes more reliable the objective measurement in real time of the relevance of the sequence. video is common for the community of viewers. This makes it possible to manage the reading module 121 more finely by the aggregation system 1 of video content according to the invention and ultimately contributes to a more effective control of the video broadcast on demand.

Preferably, the first threshold σι and the second threshold σ2 are io identical to each other: σι = σ ₂ = σ.

This arrangement contributes to simplifying the aggregation system 1 of video content according to the invention.

Preferably, the vote module 122 is designed to continue V10, when the second result β ₂ current is below the second predetermined threshold σ ₂ i 5, the continuous counting of the votes v during the playback of the video sequence a current up to 'at the end of the video sequence a, current.

Thanks to this arrangement, it is possible to follow (under certain conditions described above) in real time a representative score of the 0 relevance of the video sequence a common to the community of spectators until the end of the reading of the video sequence is current. This makes it possible to more finely manage the reading module 121 by the aggregation system 1 of video content according to the invention and ultimately contributes to the more effective control of the video broadcast on demand.

Preferably, the read module 121 is designed to adopt L2, on receipt of the first warning signal φι of the voting module 122, during the predetermined time interval Θ, an alert state ε. Under these conditions, the man-machine interfaces 1 10 are adapted to display H5 the alert state ε of the reading module 121 together with the vote session S, current. In addition, reading by the reading module 121 of the video or current sequence, submitted to the vote and the counting of the corresponding votes v by the voting module 122, continues during the voting session S, current regardless of the state warning ε adopted by the reading module 121.

Thanks to the alert state ε visible on the human-machine interfaces 1 10 (on receipt by the reading module 121 of the first warning signal φ from the voting module 122), each spectator of the community of spectators using the user terminals 1 1 is informed in real time of a particular appreciation reserved for the video sequence, current by the community of spectators. Thus, by noting the state of alert ε visible on the man-machine interfaces 1 10, a spectator, who until then, for example, hesitated to vote, can do so. As a result, the alert state ε (visible on the human-machine interfaces 1 10) expands a range of artificial intelligence instruments set up by the aggregation system 1 of video content according to the invention, this artificial intelligence ultimately contributing to the more effective control of video streaming on demand.

In this variant embodiment, the alert state ε of the read module 121 is deleted at the end of the predetermined time interval Θ, in the absence of the second warning signal φ ₂ of the vote module 122.

The selective suppression of the alert state ε contributes to widening the range of artificial intelligence instruments set up by the aggregation system 1 of video content according to the invention, which thus has functional possibilities. expanded.

Preferably, the reading module 121 is designed to close L3 the current voting session S and stop L4 the reading of the video sequence a, current subject to vote upon receipt of the second warning signal φ ₂ of the voting module 122 . Thanks to this arrangement, it is possible to shorten the reading of the current video sequence which has generated two unsatisfactory successive results when the first result βι and the second result β ₂ exceed the first threshold σι and the second threshold σ ₂ respectively. Thus, the spectators do not waste their time to visualize the video sequences deemed of no interest to the community of spectators. This makes the aggregation system 1 of video content according to the invention more user-friendly.

In this variant embodiment, the reading module 121 is designed to start L5, once the voting session S, current close, the reading of a video sequence α, + ι next from the stream of video sequences. Under these conditions, the vote module 122 is designed to be reset V1 1 at the beginning of the reading of each subsequent video sequence αι + ι.

Thanks to this arrangement, the voting module 122 is adapted to turn "in loop" according to the operating state of the reading module 121: every time the current voting session If is closed, a new polling session SM (associated with the next video sequence α, + ι) is open. This helps to exclude a priori any confusion between votes v issued H4 during the two successive voting sessions Si, Si + i, which ultimately makes the aggregation system 1 of video content according to the invention more reliable. Preferably, the voting module 122 is designed to have V12 a plurality of statuses attributable to user terminals 1 1, these statuses being different from each other, said plurality of statuses being previously stored by the voting module 122. In these conditions, the first rule Ri predetermined is designed to: · allocate V 3 to each status a voting multiplier coefficient of its own, and

• assign V14 to each user terminal 1 1 a status of its own.

Thanks to this arrangement, it is possible to individualize more the contribution of each spectator in the votes v emitted H4 by the man-machine interfaces 0 during the current session Si. For example, for each current video sequence, the spectators having a privileged status of member of the social network housing the aggregation system 1 of video content according to the invention may have more weight in terms of votes v thanks to the multiplier coefficient specific to this privileged status than other spectators do not benefit from this privileged status. As a result, the number of votes v issued H4 from the user terminals 1 1 and the number of votes v counted V2 by the voting module 122 may be different from each other. Thus, the plurality of statuses attributable to user terminals 11 expands the range of artificial intelligence instruments implemented by the video content aggregation system 1 according to the invention, which therefore has functional possibilities. expanded.

In this variant embodiment, said allocation V14 to each user terminal 11 of its own status takes place when this user terminal 11 is connected to the aggregator 12.

This arrangement makes the aggregation system 1 of video content according to the invention more reliable.

Preferably, the current polling session S has two suggestions, one of zapping the current video clip ai, the other of keeping the current video clip play M. Under these conditions, each vote v emitted H4 during the voting session S, current by the human-machine interfaces 1 10 from the user terminals 1 1 to the aggregator 12 is to confirm H6 one or the Another of these two suggestions Z, M (in Figure 1, the arrow H6 points in the blackened area around the letter Z: thus, is schematically illustrated a vote v issued in favor of M maintenance of the video sequence, current).

Thanks to this arrangement, the spectators can express their assessments in a binary way: • either by asserting the relevance of the current video sequence <¾ by choosing to maintain M its reading,

• either by expressing their disavowal vis-à-vis the current video sequence ai by choosing to zap Z the video sequence a current. This binary way of expressing the opinion makes it possible to avoid, a priori, any ambiguity as to the appreciation truly experienced by each spectator with respect to the current video sequence. This helps to make the aggregation system 1 of video content according to the invention more reliable.

In this embodiment of the invention, the second predetermined rule R2 consists in defining V15 the first selective moment xi of the current vote session S as a function of at least a predetermined number η of the voted user terminals 1 1 having at least one identical status, this predetermined number η being previously memorized by the voting module 22.

Thanks to this arrangement, the artificial intelligence of the aggregation system 1 of video content according to the invention can, for example, delay (or, on the contrary, accelerate) the arrival of the first selective moment τι of the voting session Si current if the predetermined number η voted user terminals 1 1 having at least one identical status (for example, the membership status of the social network housing video content aggregation system 1 according to the invention, as mentioned above). above) is lower (or higher) than a previously defined value (for example, 10, or 100, or 1000). Thus, the operation of the aggregation system 1 of video content according to the invention can be adjusted in real time according to the synergistic uncertainties induced by both the number of user terminals 1 1 and the status of the spectators. behind these user terminals 11. This extends the functional possibilities of the aggregation system 1 of video content according to the invention. In particular, this makes it possible to ensure greater reliability of the first result βι which sits not only on a statistically significant number "as such" of user terminals 1 1 connected to the aggregator 12 (the term "statistically significant" referring here to the probability law, called "Student t law", as described, for example, here: http: //fr.wikipedia.org/wiki/Loi Student, and more exactly on the table of the values of the quantile, called "Student's table") but also on "qualities" different from the spectators (who are behind these terminals) users 1 1) vis-à-vis their respective notorieties in the community of spectators (this notoriety being translated into language "machine" using "status" attributable to each user terminal 1 1).

In this variant embodiment, the first result βι current consists in ratio βι = VZI / VMI, where:

• the votes vzi to "zapper Z the video sequence a, current" are counted at the first moment τι selective of the voting session S, current, and

• The VMI votes to "maintain M the current video sequence" are counted at the first selective moment of the current voting session S.

Under these conditions, the first predetermined threshold σι is preferably equal to 1.

This arrangement contributes to simplifying the aggregation system 1 of video content according to the invention and to making said aggregation system 1 more user-friendly for viewers who use the user terminals 1 1. In this variant embodiment, the second result βΣ current consists in ratio 2 = vzzlv, where:

The votes vz2 for "zapping the video sequence a, current" are counted at the end of the predetermined time interval Θ of the current voting session Si, and • The VM2 votes to "maintain M the current video sequence" are counted at the end of the predetermined time interval Θ of the current voting session S.

Under these conditions, the second predetermined threshold 02 is preferably equal to 1.

This arrangement contributes to simplify the aggregation system 1 of video content according to the invention and to make said aggregation system 1 more user-friendly for viewers who use the user terminals 1 1. Preferably, the predetermined time interval Θ is equal to or less than 20 seconds: Θ <20 s.

Thanks to this arrangement, the spectators have enough time to vote during the predetermined time interval Θ. This contributes to increasing the reliability of the aggregation system 1 of video content according to the invention.

Preferably, the alert state ε of the reading module 121 displayed on the human-machine interfaces 1 10 of the user terminals 1 1 has a countdown Ξ, this countdown≡ being adapted to indicate the time remaining until the end of the predetermined time interval Θ. Thanks to the countdown Ξ, each spectator can know exactly the time remaining until the end of the predetermined time interval pour to emit H4 his vote v in time. This makes the aggregation system 1 of video content according to the invention more user-friendly.

Preferably, the voting module 122 is designed to associate V16 each current voting session S with a vote counter Σ, counted specific to said current voting session S, the human-machine interfaces 1 being adapted to display H7. the vote counterΣ, together with the corresponding current voting session S (in FIG. 1, the vote counter Σ, displays, by way of illustration, the moment when the number of votes counted VM in favor of the maintenance M of the video sequence a, current is equal to 59: v _M = 59, and the number of votes counted vz in favor of Z-zapping of the video sequence a, current is equal to 39 : v _z = 39). Thanks to the vote counter each spectator can know in real time the votes v issued H4 by the other spectators and counted during the current voting session S. This knowledge can lead, for example, to a hesitant spectator (who has not yet been voted), to vote to try to modify an instant vote result (which emerges from the vote counter, at a given moment) or a tendency to change. of voting (which emerges from the temporal evolution of a plurality of successive instantaneous voting results). Thus, the vote counter Σ expands the range of artificial intelligence instruments set up by the aggregation system 1 of video content according to the invention, which therefore has expanded functional possibilities.

In addition, the voting module 122 may be designed to synchronize V17 cyclically the vote counter, for the duration of the current voting session S, the duration of the cycle Ω between two successive synchronizations of said vote counter. , being previously memorized by the voting module 122.

Thanks to this arrangement, each spectator can see how he impacts, by his own vote, the instant vote result displayed by the vote counter. This makes the aggregation system 1 of video content according to the invention more user-friendly. In this variant embodiment, the duration of the cycle Ω between two successive synchronizations of said counter is preferably less than 10 seconds: Ω <10 s.

Thanks to this arrangement, the update of the vote counter, which takes place during its synchronization, is almost instantaneous. So, the The viewer can see almost immediately the impact that his / her own vote produces on the current voting result of the entire community of viewers, this current voting result being displayed on the vote counter Σi. Consequently, the cycle time Ω between two successive synchronizations of less than 10 seconds contributes to making the aggregation system 1 of video content according to the invention more user-friendly.

Preferably, the translation module 120 is designed to prioritize T2 the orders p emitted by the man-machine interfaces 0 of user terminals 1 1 in P requests for video content to the remote server 10, this hierarchization s' operating according to a predetermined third rule R ₃ previously stored by the translation module 120.

Thanks to this arrangement, it is possible to program the video content upstream in the stream of video sequences a (this stream being generated F1 in

1 5 responses to requests P) from the remote server 10 to the aggregator 12. This contributes to filter this downstream flow, for example, to exclude a priori from this downstream stream any video sequence that does not meet a precise charter established by the community of viewers and / or by an administrator of the video content aggregation system 1 according to the invention. The presence of this intelligent filter saves the time of the spectators who is not lost to view video sequences a priori unacceptable to the community of spectators and / or the administrator. This makes the aggregation system 1 of video content according to the invention more user-friendly.

In a variant of the invention, the third predetermined rule R ₃ consists in accepting T3 only the orders p of the human-machine interfaces 1 10 of user terminals 1 1 for video sequences whose length π is less than one. predetermined value previously stored by the translation module 120. This arrangement illustrates one of the "upstream" filters which makes it possible to deviate from the downstream stream any video sequence whose length π is greater than the predetermined value previously stored by the translation module 120. This filter makes it possible, among other things, to protect the aggregation system 1 of video content according to the invention against viral video sequences that can "loop" on themselves for several hours and thus block an expected operation of the aggregation system 1. This makes the aggregation system 1 of video content according to the invention more reliable.

As a reminder, the voting module 122 may be designed to have V12 the plurality of status attributable to user terminals 1 1. These statuses are different from each other. Said plurality of statuses is previously stored by the voting module 122. Each user terminal 1 1 connected to the aggregator 12 has a status of its own. Under these conditions, the third rule R3 predetermined may further include accepting orders p to the remote server 10 from only the user terminals 11 each having at least one privileged status previously stored by the translation module 120 selected selectively from the plurality of attributable statuses.

This arrangement illustrates another example of intelligent "upstream" filters that allow any video sequence ordered by the unauthorized viewer to be discarded from the downstream stream. By way of illustration, we can consider a community with two groups of spectators, namely: · a first group consisting only of spectators who are members of the social network who therefore have a privileged status, and

• a second group consisting only of spectators who are not members of the social network and who do not have any privileged status.

Thanks to the upstream filter described above, the orders p emitted by the non-adherent spectators belonging to the second group of the community will be hierarchized by the translation module 120 as "not acceptable" even if the length π of the corresponding video sequences is less than 20 seconds. Thus, this intelligent filter contributes to saving the time of the spectators which is not lost to view the unacceptable video sequences for the community of spectators. This makes the aggregation system 1 of video content according to the invention more user-friendly.

Preferably, the read module 121 is designed to limit L6 a list of the control and activatable functions Φ from the user terminals 1 1 via the man-machine interfaces 1 10 during the reading of the video sequence a current list of activatable control and control functions Φ being previously stored by the reading module 121.

This arrangement contributes to simplifying the man-machine interfaces 1 to, for example, facilitate handling of these interfaces by

1 5 spectators with reduced mobility and children. In addition, the smaller the list of the controllable control and control functions Φ, the less the user terminals 1 1 consume energy which makes them more efficient, especially when they are integrated in embedded devices. for example, in portable smart phones, tablets, laptops, including those embedded in vehicles, etc.

Preferably, the list of activatable control and control functions Φ comprises the following functions:

Control of the dimensions of the display field of the current video sequence of the "full screen" or "reduced screen" type on the corresponding human-machine interfaces 1 10,

Control of the resolution of the display field of current video sequences of the "high resolution resolution" or "low definition resolution" type on the corresponding human-machine interfaces 1 10, and • control of the sound volume of video sequences a, common on human-machine interfaces 1 10.

This helps to simplify the aggregation system 1 of video content according to the invention.

Preferably, the aggregator 12 furthermore comprises a video content storage module 123 linked to the reading module 121 and adapted to save (K1) the video sequences a resulting from the video content streams coming from the remote server 10 beforehand. their reading one after the other by the reading module 121.

Thanks to this arrangement, it is possible to save K1 the video sequences to be read by the reading module 121 in the storage module 123. Thus, the reading module 121 can continue to operate "locally", drawing on the video sequences has since the reading module 121, when, for example, the telecommunications network 2 between

The aggregator 12 and the remote server 10 are interrupted. Therefore, the video content storage module 123 contributes to making the aggregation system 1 of video content according to the invention more reliable.

Preferably, the vote module 122 is designed to memorize V19 the final results θ respective representative votes v counted at the end of the respective readings of the video sequences read by the reading module 121 for a lapse of time predetermined δ previously stored by the voting module 22.

Thanks to this arrangement, it is possible to identify the video sequences that have obtained, for example, the best or the worst appreciations5 in terms of the final results θ of the votes v counted at the end of the respective readings. This information can be subsequently used, for example, to selectively correct (upward or downward) the present statuses (which can be organized into several grades / skill levels) of the user terminals 1 1 that have been used to order the best and the worst video footage a. The possibility of being rewarded or, on the contrary, to be penalized by the community via said final results of the votes v counted, stimulates each spectator to act even more selectively during the preparation of the orders p to the aggregator 12, which in fine and In the long run, it contributes to reducing the number of video sequences that are of no interest to the community read by the reading module 121. This extends the range of artificial intelligence instruments implemented by the aggregation system 1 of video content according to the invention, this artificial intelligence ultimately contributing to the more effective control of video streaming on demand. In this variant embodiment, the predetermined time lapse corresponds to at least one of the following lapses of time: (a) one day; (b) a week; (c) one month; (d) one quarter; (e) one semester; (f) one year.

This extends the range of artificial intelligence instruments implemented by the aggregation system 1 of video content according to the invention.

Similarly, in this variant embodiment, the man-machine interfaces 1 10 of user terminals 1 1 are designed to display H8 each current video sequence ai with said respective final results θ representative of the votes v counted at the end of the readings. respective video sequences read by the reading module 121 during the lapse of time δ predetermined 0 previously stored by the voting module 122.

Such a display H8 of said final results θ in the form, for example, of a summary report of the three best video sequences at each lapse of time, makes the aggregation system 1 of video content according to the invention more user-friendly and informative. for the spectators.

Preferably, the vote module 122 is designed to prescribe V18 an incompressible buffer time λ between the beginning of the reading L1 of the video sequence ai by the reading module 121 and the association V1 of the voting session S, to said video sequence ai. The incompressible buffer time λ allows the spectators to appreciate the video sequence a, common before the opening of the voting session Si to said video sequence α ,. Thus, the votes v are emitted H4 by the spectators via the human-machine interfaces 10 "knowingly". This prevents the audience from voting in a hasty and / or tendentious way during, for example, the very first moments of the reading of the video sequence a, common, that is to say, when it is still impossible to make a objective opinion regarding the relevance of the current video sequence. Thus, the presence of the incompressible buffer time λ expands the range of artificial intelligence instruments set up by the aggregation system 1 of video content according to the invention, this artificial intelligence ultimately contributing to the more effective control of the video streaming on demand.

Preferably, for each current video sequence having a length π ,, the incompressible buffer time λ between the beginning of the reading L1 of the video sequence ai by the reading module 121 and the association V1 of the voting session S, to said video sequence satisfies the following inequality: 0.1 ^* π, <λ <0A ^* m.

This arrangement introduces the dynamic incompressible λ buffer time, that is to say, depending on the length m specific to the video sequence ¾. This makes the artificial intelligence of the aggregation system 1 of video content according to the invention more flexible.

Preferably, the incompressible buffer time λ between the beginning of the reading L1 of the video sequence a, by the reading module 121 and the association V1 of the voting session S, with said video sequence ai, is greater than 10. seconds: λ> 10 s.

This arrangement introduces the static incompressible λ buffer time, that is to say, independent of the length π, specific to the video sequence. This makes the aggregation system 1 of video content according to the invention simpler.

A synergy between these two aspects (static and dynamic) of the incompressible buffer time λ is obtained when the latter responds to two inequalities at a time: 0.1 * π <λ <0.4 * π, and λ> 10 ε.

This arrangement is particularly relevant for the short-lived video sequences, that is to say, for the current a video sequences having the lengths π, between 1 1 seconds and 25 seconds. Thus, said synergy between the two aspects (static and dynamic) of the incompressible buffer time λ contributes to ensuring an objective assessment by the community, via the reasoned votes (because allowed only after more than 10 seconds of reflection), even for video sequences with short currents, which ultimately widens the functional possibilities of the

Video content aggregation system 1 according to the invention.

Preferably, each user terminal 11 is designed to transmit H9, during each voting session If, via its man-machine interface 1 10, at least two successive votes v of appreciation of the current video sequence ai to the aggregator 12, these two successive votes v being spaced from each other by a predetermined reflection time Γ previously recorded by the user terminal 1.

Thanks to this arrangement, each spectator can correct, after the predetermined reflection time,, his first vote v (for example, to zap Z the video sequence a, current) transmitted, for example, before the display 5 of the state alarm ε, by its second vote v (for example, to maintain M the video sequence a, current), for example, after the display of the alert state ε. This helps to ultimately obtain the opinion of the community of spectators even more objective vis-à-vis the relevance of the video sequence a common. Thus, this possibility of issuing at least two votes v successive spaced apart from each other by the predetermined reflection time Γ expands the range of artificial intelligence instruments set up by the aggregation system 1 of video content according to the invention, this artificial intelligence ultimately contributing more effective mastery of video streaming on demand.

In this embodiment, for each video sequence ¾ current having a length π ,, the predetermined reflection time Γ between the two successive votes v issued by the human-machine interfaces 1 10 to the aggregator 12, satisfies the following inequation: 0.1 * 7ii <Γ <0.4 * πι.

This arrangement introduces the dynamic reflection time c'est, that is to say, depending on the length n specific to the video sequence. This makes the artificial intelligence of the aggregation system 1 of video content according to the invention more flexible. Preferably, the reflection time Γ is greater than 10 seconds: r> 10 s.

This arrangement introduces the reflection time Γ static, that is to say, independent of the length π, specific to the video sequence. This makes the aggregation system 1 of video content according to the invention simpler. A synergy between these two aspects (static and dynamic) of the reflection time Γ is obtained when the latter responds to two inequalities at a time: 0.1 * π, <Γ <0.4 * π, and Γ> 10 s.

This arrangement is particularly relevant for short current video sequences, i.e., for common video sequences having lengths m between 1 second and 25 seconds. Thus, said synergy between the two aspects (static and dynamic) of the reflection time Γ contributes to ensuring an objective appreciation by the community, via the reasoned votes (because allowed only after more than 10 seconds of reflection), even video sequences. , short run, which ultimately widens the functional possibilities of the aggregation system 1 of video content according to the invention.

Preferably, the translation module 120 is furthermore designed to affiliate T4 with each previously prioritized request P according to the third predetermined rule R ₃ , with a particular first callback Yi intended for the user terminal 1 1, the sender of the order p corresponding to this request P, the human-machine interface 110 of said user terminal 1 1 (issuer of the order p) being adapted to display H10 this particular first reminder

Thanks to this arrangement, it is possible to bring to the attention of the viewer concerned (that is to say, to the one who sent, via his user terminal 11, the order p corresponding to this hierarchical request P) privileged information, for example:

• privileged information confirming the acceptance or refusal of the issued order,

• a privileged information with an estimate of the moment of the reading of the video sequence, etc.

This makes the aggregation system 1 of video content according to the invention more user-friendly.

Preferably, the voting module 122 is designed to affiliate V20 with each final result θ (which is representative of the votes v counted at the end of the reading of each video sequence a, current, that is to say, at the closure of each voting session If current) with a second particular reminder Y2 to the user terminal 1 1, issuer of the order p aimed at said said current video sequence <¾ (corresponding to said voting session S, current closed) ), the human-machine interface 110 of said user terminal 1 1 (emitter of the order p) being adapted to display H1 1 this particular second reminder Y ₂ . Thanks to this arrangement, it is possible to bring to the attention of the viewer concerned a privileged piece of information, for example:

• privileged information confirming the final result θ representative of the votes v counted for the video sequence a, orderly current

5 previously by this concerned spectator (via the order p sent by the user terminal 1 1 of this viewer concerned),

• privileged information aimed at trophies (or penalties) intended for this spectator concerned as a function of the final result θ obtained.

In an alternative manner to the embodiment variant illustrated in FIG. 1 and described above with the current voting session S, comprising two suggestions, one of zapping the current video sequence Z, the other consisting in maintaining M the video sequence a, common,

The current voting session S may consist of a single suggestion to zap the current video sequence a (not shown in FIG. 1). Under these conditions, each vote v emitted H4 during the voting session S, current by the man-machine interfaces 110 from the user terminals 1 1 to the aggregator 12 consists in confirming this unique suggestion of zapping the sequence Z video a, current.

This arrangement simplifies the aggregation system 1 of video content according to the invention to better adapt its operation to the reduced analysis capabilities specific to certain categories of spectators (for example, to children). Another possible alternative is for the current voting session S, which may consist of a single suggestion to keep the current video sequence M (variant not shown in FIG. 1). Under these conditions, each vote v emitted H4 during the voting session S, current by the human-machine interfaces 1 10 from the user terminals 1 1 to the aggregator 12 consists in confirming this unique suggestion to maintain M the current video sequence.

This arrangement simplifies the aggregation system 1 of video content according to the invention to better adapt its operation to the reduced analysis capabilities specific to certain categories of spectators (for example, to children).

Claims

claims

1. Video content aggregation system (1) comprising:

A remote server (10) comprising a database (100) with video content consisting of video sequences (a), user terminals (1 1) connected to each other and to the remote server (10). ) via a telecommunication network (2) via an aggregator (12) of video content, the user terminals (1 1) each comprising a human-machine interface (1 10) adapted to transmit (H 1 ) orders (p) to the aggregator (12) and to display (H2) the video sequences read by the aggregator (12),

the aggregator (12) comprising a translation module (120) adapted to translate (T1) the orders (p) sent by the human-machine interfaces (1 10) of user terminals (1 1) into requests (P) video content to the remote server (10),

the remote server (10) being adapted to generate (F1) a stream of video sequences (a) to the aggregator (12) in response to its requests (P) for video content,

the aggregator (12) further comprising a read module (121) of video content adapted to read (L1), one after the other, the video sequences (a) from the stream of video sequences,

characterized in that the aggregator (12) further comprises a voting module (122) linked with the reading module (121) and adapted to:

Associating (V1) each video sequence (,) currently being read by the reading module (121) with a current vote session (Si) specific to said current video sequence (α,), the human-machine interfaces (1). 10) user terminals (1 1) being adapted to: o display (H3) the current voting session (S,) together with the corresponding current video sequence (ai), o transmit (H4) votes (v) to the aggregator (12) during the voting session ( If) current,

5 · continuously counting (V2) the votes (v) sent (H4) by the human-machine interfaces (1 10) to the aggregator (12), this continuous counting of votes (v) taking place in real time without interruption, by adding them to each other, as and when the votes (v) in the voting module (122), according to a first predetermined rule (Ri) previously stored by the voting module

(1 22),

• obtain (V3) a first result (βι) current representative of the votes (v) counted at a first moment (τι) selective of the current voting session (S,), the choice of the first (n) selective moment taking place according to a predetermined second rule (R2) previously stored by the voting module (122),

• compare (V4) this first result (βι) current of the current voting session (S,) with a predetermined first threshold (σι) previously stored by the voting module (122), 0 · supply (V5) a first signal warning (φι) to the read module (121) when the first result (βι) current exceeds the first threshold (σι) predetermined.

Video content aggregation system (1) according to claim 1, characterized in that the voting module (122) is designed to: · continue (V6) the continuous counting of votes (v) during an interval a predetermined time (Θ) calculated from the first selective moment (τι), said predetermined time interval (Θ) being previously stored by the voting module (122), Obtaining (V7) a second result (β ₂ ) representative of the votes (v) counted at the end of the predetermined time interval (Θ) of the current voting session (S,), the votes (v) counted during the predetermined time interval (Θ) in addition to those obtained at the first (τι) selective time,

Comparing (V8) this second current result (βΣ) with a second predetermined threshold (02) different from the predetermined first threshold (σι),

• supply (V9) a second warning signal (92) to the reading module (121) when the second result (β2) current exceeds the second threshold (σ ₂ ) predetermined.

Video content aggregation system (1) according to claim 2, characterized in that the voting module (122) is designed to continue (V10), when the second result (βΣ) current is below the second threshold. (σ ₂ ) predetermined, the continuous counting of votes (v) during the playback of the current video sequence (¾) to the end of the current video sequence (,).

Video content aggregation system (1) according to claim 2 or 3, characterized in that the reading module (121) is adapted to adopt (L2) on receipt of the first warning signal (φι) of the polling module (122), during the predetermined time interval (Θ), an alert state (ε), the man-machine interfaces (1 10) being adapted to display (H5) the alert state ( ε) of the reading module (121) together with the current voting session (S,), reading by the reading module (121) of the current video sequence (α,) submitted to the vote and the counting of the votes (v ) corresponding by the voting module (122) continuing during the current voting session (Si) regardless of the alert state (ε) adopted by the reading module (121).

5. System for aggregating (1) video content according to claim 4, characterized in that the alert state (ε) of the reading module (121) is deleted at the end of the predetermined time interval (Θ), in the absence of the second warning signal (φ ₂ ) of the voting module (122).

Video content aggregation system (1) according to one of claims 2 to 5, characterized in that the read module (121)

5 is designed to close (L3) the current voting session (Si) and to stop (L4) the reading of the current video sequence (a ,,) submitted to vote on receipt of the second alert signal (φ ₂ ) of the module of vote (122).

Video content aggregation system (1) according to claim 6, characterized in that the reading module (121) is designed to start at (L5), once the current voting session (Si) is closed, playing a next video sequence (oti + i) from the stream of video sequences, and in that the voting module (122) is designed to be reset (V1 1) at the beginning of the playback of each subsequent video sequence (α, + ι).

An aggregation system (1) of video content according to any of claims 1 to 7, characterized in that the voting module (122) is adapted to arrange (V12) a plurality of the statuses attributable to the terminals users (1 1), these statuses being different from each other, said plurality of statuses being previously memorized by the voting module (122), and in that the first predetermined rule (Ri) is designed for:

• allocate (V13) to each status a voting multiplier coefficient of its own, and

• assign (V14) to each user terminal (1 1) a status of its own. 5

Video content aggregation system (1) according to one of claims 1 to 8, characterized in that the current voting session (Si) comprises two suggestions, one of zapping (Z) the sequence video (α,) current, the other consisting in maintaining (M) the reading of the current video sequence (α,), and in that each vote (v) transmitted (H4) during the current polling session (S,) by the human-machine interfaces (1 10) from the user terminals (1 1) to the aggregator (12) is to confirm (H6) one or the other other of these two suggestions (Z), (M).

An aggregation system (1) of video content according to the claim

9, characterized in that the second predetermined rule (R2) consists in defining (V15) the first (n) selective moment of the current voting session (Si) as a function of at least a predetermined number (η) of the terminals Voted users (1 1) having at least one identical status, this predetermined number (η) being previously memorized by the voting module (122).

1 1. Video content aggregation system (1) according to one of Claims 4 to 10, characterized in that the alert state (ε) of the read module (121) displayed on the man-machine interfaces

I 5 (1 10) user terminals (1 1) comprises a countdown (Ξ), this countdown (Ξ) being adapted to indicate the time remaining until the end of the time interval ( Θ) predetermined.

Video content aggregation system (1) according to one of claims 1 to 11, characterized in that the 0 vote module (122) is designed to:

Associating (V16) each current voting session (S,) with a counting counter (Σ,) counted specific to said current voting session (Si), the human-machine interfaces (1 10) being adapted to display (H7 ) the vote counter (Σ,) together with the corresponding current vote session (S,) 5,

• synchronize (V17) cyclically the vote counter (Σ,) for the duration of the current voting session (S,), the cycle time (Ω) between two successive synchronizations of said vote counter (Σ,) being previously memorized by the voting module (122).

Video content aggregation system (1) according to one of claims 1 to 12, characterized in that the translation module (120) is designed to prioritize (T2) the commands (p) issued by the interfaces. human-machine (1 10) of user terminals (1 1) in requests (P) for video content to the remote server (10), this prioritization operating according to a predetermined third rule (R ₃ ) previously stored by the translation module (120).

Video content aggregation system (1) according to claim 13, characterized in that the predetermined third rule (R) consists in accepting (T3) only the commands (p) of the human-machine interfaces (1 10) of user terminals (11) for video sequences (a) whose length (π) is less than a predetermined value previously stored by the translation module (120).

Video content aggregation system (1) according to one of claims 1 to 14, characterized in that the read module (121) is adapted to limit (L6) a list of the control and control functions. activatable (Φ) from the user terminals (1 1) via the human-machine interfaces (10) during the reading of the current video sequence (α,), a list of the activatable control and control functions (Φ) being previously memorized by the reading module (121).