Multi-person video conference picture layout method based on audio analysis

Abstract

The invention discloses a multi-person video conference picture layout method based on audio analysis. The method comprises the following steps: S1, assigning initial values to the audio of each participant, wherein the initial values comprise a participant weight, a speech weight and a window weight; s2, calculating voice energy of each path of audio; s3, calculating to obtain a real-time voice coefficient of each path of audio; s4, the conference participation weight of each path of audio is updated; s5, updating the window weight of each path of audio; s6, updating the main speaker; s7, updating the size coefficient of each video picture; s8, updating the position coefficient of each video picture; s9, calculating to obtain the actual size and position condition of each current picture, namely obtaining a final picture layout; according to the invention, the conditions of a plurality of speakers in a conference can be clearly and visually displayed, so that participants can better pay attention to the conditions of the speakers in the conference.

Description

Technical Field

[0001] The present invention relates to the field of digital audio technology, and in particular to a multi-person video conference screen layout method based on audio analysis. Background Art

[0002] At present, in conference room environments, conference tablets are often used for online meetings. Conference tablets are smart tablets that integrate functions such as smart writing, wireless screen transmission, video conferencing, and remote collaboration. Conference tablets can be installed with video conferencing software to conduct video conferences with other online users. In existing technical solutions, the screen layout is mainly the following: 1. Depending on who is currently speaking, the current speaker will be placed on the main screen, and other non-speaking users will be displayed as small pictures on the edge; 2. All participants are displayed on a single screen, and the current speaker is identified by drawing a box outside the screen.

[0003] However, when there are many participants, for scenarios where multiple people are discussing with each other, the above-mentioned layout method cannot intuitively and clearly display the multiple speakers closest to them. In a complex discussion environment, especially when multiple participants speak almost at the same time, this simple identification method may not be enough to quickly attract the attention of all participants, thereby missing key information. It lacks the flexibility to dynamically adjust the screen according to the actual situation of the meeting and cannot effectively support diverse forms of conference interaction.

[0004] Therefore, it is necessary to propose a new multi-person video conference screen layout method, which can clearly and intuitively display the situation of multiple speakers in the conference, so that participants can better pay attention to the situation of the speakers in the conference. Summary of the invention

[0005] The present invention provides a multi-person video conference screen layout method based on audio analysis, comprising the following steps:

[0006] S1. Assign initial values ​​to the audio of each participant: participant weight, speech weight and window weight;

[0007] S2, calculating the speech energy of each audio channel;

[0008] S3, calculating the real-time speech coefficient of each audio channel through the speech energy sequence of each audio channel within a period of time;

[0009] S4. Update the conference weight of each audio channel according to the current voice coefficient of each audio channel in the conference venue;

[0010] S5. Update the window weight of each audio channel according to the current voice coefficient of each audio channel in the conference room;

[0011] S6. Update the main speaker based on the current audio participant weights and window weights of each channel;

[0012] S7. Based on the current audio conference weights, voice coefficients, and window weights of each channel, update the size coefficients of each video screen;

[0013] S8. Based on the current weights of the audio windows of each channel, update the position coefficient of each video screen;

[0014] S9. According to the current main speaker, the size coefficients of each screen, and the sorting coefficients of each screen, the actual size and position of each screen are calculated to obtain the final screen layout.

[0015] Furthermore, speech energy represents the instantaneous intensity of sound energy in a frame of audio signal in an audio stream;

[0016] Audio PCM time domain uses data x[i];

[0017] Among them, x[i] is the time domain amplitude sampling sequence;

[0018] Let the current audio energy be Pv, then:

[0019] Pv = SUM(X[i]^2);

[0020] Here, i takes values ​​in the closed interval [a,b].

[0021] Furthermore, the speech coefficient represents the speech intensity of the speaker of the audio stream in a short period of time recently;

[0022] The audio frame interval is recorded as T;

[0023] Let Pv[i] be the speech energy at the time T*i from the current moment, that is:

[0024] The current latest speech energy is Pv[0];

[0025] The speech energy of the previous frame is Pv[1];

[0026] The speech energy of the first two frames is Pv[2], and so on;

[0027] Let the current voice coefficient be Vp, then:

[0028] Vp=SUM(Pv[i]*(1 / (1+i / 5)));

[0029] i is all the values ​​that make T*i<1 second.

[0030] Furthermore, the conference participation weight represents the voice weight of one audio channel in the current conference;

[0031] At the current moment, there are N audio channels in the conference room, and the speech coefficient of the kth audio channel is Vp[k];

[0032] For the kth audio channel:

[0033] The weight of the participant at the current frame is M[k][i];

[0034] The weight of the participant at the previous frame is recorded as M[k][i-1];

[0035] Let the initial participation weight be M[k][0]=1;

[0036] The sum of the voice coefficients of all N audio channels is Vps, then:

[0037] Vps = SUM(Vp[k]), k belongs to [0, N];

[0038] The current conference gain of the k-th audio channel is A[k], and:

[0039] Am[k]=(Vp[k] / Vps>30%)? Vp[k]:0;

[0040] The current conference weight is calculated based on the current conference gain and the previous frame conference weight:

[0041] M[k][i]=0.8*M[k][i-1]+Am[k].

[0042] Furthermore, the window weight is a sequence;

[0043] The audio frame interval is recorded as T;

[0044] The sequence length is L, let L = 60 seconds / T;

[0045] For the k-th audio channel, the window weight is W[k], where the i-th element is W[k][i];

[0046] The value range of i is [0, L-1];

[0047] For the k-th audio channel, Vp[k][i] is the speech energy at the time T*i from the current moment, that is:

[0048] The current latest voice coefficient is Vp[k][0];

[0049] The speech coefficient of the previous frame is Vp[k][1];

[0050] The speech coefficient of the first two frames is Vp[k][2], and so on;

[0051] Then, let W[k][i]=Pv[k][i].

[0052] Furthermore, for the main speaker coefficient and the main speaker update:

[0053] For the k-th audio channel, the main speaker coefficient of the current frame is recorded as S[k];

[0054] The audio frame interval is recorded as T;

[0055] Let S[k]=3*M[k]+SUM(W[k][j]);

[0056] SUM is the sum of W[k][j] over the range of j, where the value of j is in the range of [0,15 seconds / T];

[0057] M[k] is the conference weight of the k-th audio channel at the current moment;

[0058] The main speaker coefficient of each user is compared, and the user with the largest coefficient is the current main speaker.

[0059] Furthermore, for the screen size coefficient update:

[0060] The size factor determines whether each screen will be displayed and at what size;

[0061] The weight of the participant at the current frame is M[k][i];

[0062] For the k-th audio channel, the size coefficient of the current frame is recorded as Q[k][i], and the size coefficient of the previous frame is recorded as Q[k][i-1];

[0063] Let the size coefficient at the initial moment be Q[k][0]=0;

[0064] The sum of all elements of the audio window weight of the k-th audio channel is recorded as W[k];

[0065] The sum of W[k] of all audio channels is Ws;

[0066] If W[k] of the k-th audio channel is not in the top five, set Q[k][i] = 0;

[0067] If W[k] / Ws of the k-th audio channel is less than 10%, set Q[k][i]=0;

[0068] Other cases:

[0069] If Q[k][i-1]=0, then Q[k][i]=100;

[0070] Otherwise, Q[k][i]=Q[k][i-1]*(1+W[k] / Ws).

[0071] Furthermore, for the picture sorting coefficient update:

[0072] The sorting coefficient determines the rules by which each picture is sorted;

[0073] The serial number of the participant whose current size coefficient is not 0 is recorded as k;

[0074] Calculate the sorting coefficient P[k] for the k-th audio channel;

[0075] The window weight of the current audio channel is W[k], where the value of each element is W[k][j];

[0076] Sorting coefficient P[k]=Q[k]>0? M[k]+SUM(W[k][j]):0;

[0077] Q[k] is the size coefficient of the k-th audio channel at the current moment;

[0078] SUM is the sum of W[k][j] over the range of j, where the value of j is in the range of [0,30 seconds / T];

[0079] M[k] is the participation weight of the k-th audio channel at the current moment.

[0080] Furthermore, for the picture size calculation:

[0081] The total width of the current screen is WIDTH, and the total height is HEIGHT;

[0082] For a user with sequence number k, its size coefficient is Q[k] and its ranking coefficient is P[k];

[0083] Let the serial number of the current main speaker be m, let the sum of the size coefficients of all users be Qs, and let the number of users whose size coefficients are not 0 except the main speaker be r;

[0084] The display width width[k] of the screen of user number k;

[0085] If k == m, then width[k] = (0.565-0.65*r)*WIDTH;

[0086] If k! =m, then width[k]=(0.435-0.65*r)*WIDTH*Q[k] / (Qs–Q[m]);

[0087] The display height of the screen of user number k is height[k];

[0088] height[k]=(k==m||Q[k]>0)? HEIGHT:0

[0089] Furthermore, for the screen position determination:

[0090] For users with sequence number k, the ranking coefficient is P[k];

[0091] The main speaker screen is placed on the far left;

[0092] For other users with width[k]>0, they are arranged from right to left according to P[k] from large to small, starting from the right side of the main speaker.

[0093] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0094] The present invention dynamically adjusts the size and position of the video screen of each participant by analyzing audio data in real time, ensuring that important speakers obtain a larger display area, thereby improving meeting efficiency and viewing experience; for those participants who do not speak often or speak less, the system can reduce the display size of their screens or even hide their screens, thereby optimizing the overall layout of the video screen without sacrificing key information; real-time adjustment of the screen layout can instantly reflect the dynamic changes of the meeting, allowing participants to more intuitively feel the progress and focus of the meeting, and enhancing the interactivity and sense of participation in the meeting; automated and intelligent screen layout adjustment reduces the burden of manual screen adjustment on conference organizers, thereby improving the efficiency and convenience of conference management. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0096] Figure 1 A schematic diagram of a multi-person video conference screen layout method based on audio analysis. DETAILED DESCRIPTION

[0097] The technical solution of the present invention will be described below in conjunction with the accompanying drawings of the present invention, but the described embodiments are only part of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present invention.

[0098] like Figure 1 As shown, the present invention provides a multi-person video conference screen layout method based on audio analysis, comprising the following steps:

[0099] S1. Assign initial values ​​to the audio of each participant: participant weight, speech weight and window weight;

[0100] S2, calculating the speech energy of each audio channel;

[0101] S3, calculating the real-time speech coefficient of each audio channel through the speech energy sequence of each audio channel within a period of time;

[0102] S4. Update the conference weight of each audio channel according to the current voice coefficient of each audio channel in the conference venue;

[0103] S5. Update the window weight of each audio channel according to the current voice coefficient of each audio channel in the conference room;

[0104] S6. Update the main speaker based on the current audio participant weights and window weights of each channel;

[0105] S7. Based on the current audio conference weights, voice coefficients, and window weights of each channel, update the size coefficients of each video screen;

[0106] S8. Based on the current weights of the audio windows of each channel, update the position coefficient of each video screen;

[0107] S9. According to the current main speaker, the size coefficients of each screen, and the sorting coefficients of each screen, the actual size and position of each screen are calculated to obtain the final screen layout.

[0108] In this embodiment, the speech energy represents the instantaneous intensity of the sound energy in a frame of audio signal in the audio stream;

[0109] Audio PCM time domain uses data x[i];

[0110] Among them, x[i] is the time domain amplitude sampling sequence;

[0111] Let the current audio energy be Pv, then:

[0112] Pv = SUM(X[i]^2);

[0113] Here, i takes values ​​in the closed interval [a,b].

[0114] In this embodiment, the speech coefficient represents the speech intensity of the speaker of the audio stream in a recent short period of time;

[0115] The audio frame interval is recorded as T;

[0116] Let Pv[i] be the speech energy at the time T*i from the current moment, that is:

[0117] The current latest speech energy is Pv[0];

[0118] The speech energy of the previous frame is Pv[1];

[0119] The speech energy of the first two frames is Pv[2], and so on;

[0120] Let the current voice coefficient be Vp, then:

[0121] Vp=SUM(Pv[i]*(1 / (1+i / 5)));

[0122] i is all the values ​​that make T*i<1 second.

[0123] In this embodiment, the conference participation weight represents the voice weight of one audio channel in the current conference;

[0124] At the current moment, there are N audio channels in the conference room, and the speech coefficient of the kth audio channel is Vp[k];

[0125] For the kth audio channel:

[0126] The weight of the participant at the current frame is M[k][i];

[0127] The weight of the participant at the previous frame is recorded as M[k][i-1];

[0128] Let the initial participation weight be M[k][0]=1;

[0129] The sum of the voice coefficients of all N audio channels is Vps, then:

[0130] Vps = SUM(Vp[k]), k belongs to [0, N];

[0131] The current conference gain of the k-th audio channel is A[k], and:

[0132] Am[k]=(Vp[k] / Vps>30%)? Vp[k]:0;

[0133] The current conference weight is calculated based on the current conference gain and the previous frame conference weight:

[0134] M[k][i]=0.8*M[k][i-1]+Am[k].

[0135] In this embodiment, the window weight is a sequence;

[0136] The audio frame interval is recorded as T;

[0137] The sequence length is L, let L = 60 seconds / T;

[0138] For the k-th audio channel, the window weight is W[k], where the i-th element is W[k][i];

[0139] The value range of i is [0, L-1];

[0140] For the k-th audio channel, Vp[k][i] is the speech energy at the time T*i from the current moment, that is:

[0141] The current latest voice coefficient is Vp[k][0];

[0142] The speech coefficient of the previous frame is Vp[k][1];

[0143] The speech coefficient of the first two frames is Vp[k][2], and so on;

[0144] Then, let W[k][i]=Pv[k][i].

[0145] In this embodiment, the main speaker coefficient and the main speaker are updated:

[0146] For the k-th audio channel, the main speaker coefficient of the current frame is recorded as S[k];

[0147] The audio frame interval is recorded as T;

[0148] Let S[k]=3*M[k]+SUM(W[k][j]);

[0149] SUM is the sum of W[k][j] over the range of j, where the value of j is in the range of [0,15 seconds / T];

[0150] M[k] is the conference weight of the k-th audio channel at the current moment;

[0151] The main speaker coefficient of each user is compared, and the user with the largest coefficient is the current main speaker.

[0152] In this embodiment, the picture size coefficient is updated:

[0153] The size factor determines whether each screen will be displayed and at what size;

[0154] The weight of the participant at the current frame is M[k][i];

[0155] For the k-th audio channel, the size coefficient of the current frame is recorded as Q[k][i], and the size coefficient of the previous frame is recorded as Q[k][i-1];

[0156] Let the size coefficient at the initial moment be Q[k][0]=0;

[0157] The sum of all elements of the audio window weight of the k-th audio channel is recorded as W[k];

[0158] The sum of W[k] of all audio channels is Ws;

[0159] If W[k] of the k-th audio channel is not in the top five, set Q[k][i] = 0;

[0160] If W[k] / Ws of the k-th audio channel is less than 10%, set Q[k][i]=0;

[0161] Other cases:

[0162] If Q[k][i-1]=0, then Q[k][i]=100;

[0163] Otherwise, Q[k][i]=Q[k][i-1]*(1+W[k] / Ws).

[0164] In this embodiment, the picture sorting coefficient is updated:

[0165] The sorting coefficient determines the rules by which each picture is sorted;

[0166] The serial number of the participant whose current size coefficient is not 0 is recorded as k;

[0167] Calculate the sorting coefficient P[k] for the k-th audio channel;

[0168] The window weight of the current audio channel is W[k], where the value of each element is W[k][j];

[0169] Sorting coefficient P[k]=Q[k]>0? M[k]+SUM(W[k][j]):0;

[0170] Q[k] is the size coefficient of the k-th audio channel at the current moment;

[0171] SUM is the sum of W[k][j] over the range of j, where the value of j is in the range of [0,30 seconds / T];

[0172] M[k] is the participation weight of the k-th audio channel at the current moment.

[0173] In this embodiment, the picture size is calculated as follows:

[0174] The total width of the current screen is WIDTH, and the total height is HEIGHT;

[0175] For a user with sequence number k, its size coefficient is Q[k] and its ranking coefficient is P[k];

[0176] Let the serial number of the current main speaker be m, let the sum of the size coefficients of all users be Qs, and let the number of users whose size coefficients are not 0 except the main speaker be r;

[0177] The display width width[k] of the screen of user number k;

[0178] If k == m, then width[k] = (0.565-0.65*r)*WIDTH;

[0179] If k! =m, then width[k]=(0.435-0.65*r)*WIDTH*Q[k] / (Qs–Q[m]);

[0180] The display height of the screen of user number k is height[k];

[0181] height[k]=(k==m||Q[k]>0)? HEIGHT:0

[0182] In this embodiment, the screen position is determined as follows:

[0183] For users with sequence number k, the ranking coefficient is P[k];

[0184] The main speaker screen is placed on the far left;

[0185] For other users with width[k]>0, they are arranged from right to left according to their P[k] values, starting from the right side of the main speaker. That is, the user with the largest P[k] is next to the main speaker on the left, at the far left of the other users, and the user with the smallest P[k] is at the far right of the screen.

[0186] The present invention has many application scenarios, including but not limited to the following description scenarios:

[0187] In remote office scenarios, multi-person video conferencing has become an important way of daily communication. The present invention can automatically optimize the screen layout to ensure that the screen of important speakers is highlighted, thereby improving meeting efficiency.

[0188] In online education scenarios, teachers may need to interact with students frequently. The present invention can automatically adjust the screen layout of teachers and students according to the speech situation to ensure that students can clearly see the teacher's demonstration and explanation;

[0189] In telemedicine consultation, communication between doctors and patients is crucial. By automatically adjusting the screen layout, it can ensure that the key remarks of doctors or patients are highlighted, thus improving the consultation effect.

[0190] During corporate training, lecturers may need to interact and discuss frequently with trainees. The present invention can adjust the screen layout in real time to ensure that the screens of lecturers and trainees are reasonably displayed, thereby improving the training effect.

[0191] In large-scale online seminars, due to the large number of participants, manual adjustment of the screen layout becomes impractical. The present invention can automatically adjust the screen layout according to the speaking situation to ensure that the screens of important speakers are highlighted, while optimizing the overall screen layout to improve the viewing experience of participants.

Claims

1. A method for multi-person video conference screen layout based on audio analysis, characterized in that: The following steps are involved: S1. Assign initial values ​​to each participant's audio: participant weight, speech weight and window weight; S2, calculating the speech energy of each audio channel; S3, calculating the real-time speech coefficient of each audio channel through the speech energy sequence of each audio channel within a period of time; S4. Update the conference weight of each audio channel according to the current voice coefficient of each audio channel in the conference venue; S5. Update the window weight of each audio channel according to the current voice coefficient of each audio channel in the conference room; S6. Update the main speaker based on the current audio participant weights and window weights of each channel; S7. Based on the current audio conference weights, voice coefficients, and window weights of each channel, update the size coefficients of each video screen; S8. Based on the current weights of the audio windows of each channel, update the position coefficient of each video screen; S9. According to the current main speaker, the size coefficients of each screen, and the sorting coefficients of each screen, the actual size and position of each screen are calculated to obtain the final screen layout.

2. The method for multi-person video conference screen layout based on audio analysis according to claim 1 is characterized in that: Speech energy represents the instantaneous intensity of sound energy in a frame of audio signal in the audio stream; Audio PCM time domain uses data x[i]; Among them, x[i] is the time domain amplitude sampling sequence; Let the current audio energy be Pv, then: Pv = SUM(X[i]^2); Here, i takes values ​​in the closed interval [a,b].

3. The method for multi-person video conference screen layout based on audio analysis according to claim 2 is characterized in that: The speech coefficient represents the speech intensity of the speaker in the audio stream in a short period of time; The audio frame interval is recorded as T; Let Pv[i] be the speech energy at the time T*i from the current moment, that is: The current latest speech energy is Pv[0]; The speech energy of the previous frame is Pv[1]; The speech energy of the first two frames is Pv[2], and so on; Let the current voice coefficient be Vp, then: Vp=SUM(Pv[i]*(1 / (1+i / 5))); i is all the values ​​that make T*i<1 second.

4. The method for multi-person video conference screen layout based on audio analysis according to claim 3 is characterized in that: The conference weight represents the voice weight of one audio channel in the current conference. At the current moment, there are N audio channels in the conference room, and the speech coefficient of the kth audio channel is Vp[k]; For the kth audio channel: The weight of the participant at the current frame is M[k][i]; The weight of the participant at the previous frame is recorded as M[k][i-1]; Let the initial participation weight be M[k][0]=1; The sum of the voice coefficients of all N audio channels is Vps, then: Vps = SUM(Vp[k]), k belongs to [0, N]; The current conference gain of the k-th audio channel is A[k], and: Am[k]=(Vp[k] / Vps>30%)? Vp[k]:0; The current conference weight is calculated based on the current conference gain and the previous frame conference weight: M[k][i]=0.8*M[k][i-1]+Am[k].

5. The method for multi-person video conference screen layout based on audio analysis according to claim 4 is characterized in that: The window weight is a sequence; The audio frame interval is recorded as T; The sequence length is L, let L = 60 seconds / T; For the k-th audio channel, the window weight is W[k], where the i-th element is W[k][i]; The value range of i is [0, L-1]; For the k-th audio channel, Vp[k][i] is the speech energy at the time T*i from the current moment, that is: The current latest voice coefficient is Vp[k][0]; The speech coefficient of the previous frame is Vp[k][1]; The speech coefficient of the first two frames is Vp[k][2], and so on; Then, let W[k][i]=Pv[k][i].

6. The method for multi-person video conference screen layout based on audio analysis according to claim 5 is characterized in that: For main speaker coefficient and main speaker update: For the k-th audio channel, the main speaker coefficient of the current frame is recorded as S[k]; The audio frame interval is recorded as T; Let S[k]=3*M[k]+SUM(W[k][j]); SUM is the sum of W[k][j] over the range of j, where the value of j is in the range of [0,15 seconds / T]; M[k] is the conference weight of the k-th audio channel at the current moment; The main speaker coefficient of each user is compared, and the user with the largest coefficient is the current main speaker.

7. The method for multi-person video conference screen layout based on audio analysis according to claim 6 is characterized in that: For screen size factor update: The size factor determines whether each screen will be displayed and at what size; The weight of the participant at the current frame is M[k][i]; For the k-th audio channel, the size coefficient of the current frame is recorded as Q[k][i], and the size coefficient of the previous frame is recorded as Q[k][i-1]; Let the size coefficient at the initial moment be Q[k][0]=0; The sum of all elements of the audio window weight of the k-th audio channel is recorded as W[k]; The sum of W[k] of all audio channels is Ws; If W[k] of the k-th audio channel is not in the top five, set Q[k][i] = 0; If W[k] / Ws of the k-th audio channel is less than 10%, set Q[k][i]=0; Other cases: If Q[k][i-1]=0, then Q[k][i]=100; Otherwise, Q[k][i]=Q[k][i-1]*(1+W[k] / Ws).

8. The method for multi-person video conference screen layout based on audio analysis according to claim 7, characterized in that: For screen sorting coefficient update: The sorting coefficient determines the rules by which each picture is sorted; The serial number of the participant whose current size coefficient is not 0 is recorded as k; Calculate the sorting coefficient P[k] for the k-th audio channel; The window weight of the current audio channel is W[k], where the value of each element is W[k][j]; Sorting coefficient P[k]=Q[k]>0? M[k]+SUM(W[k][j]):0; Q[k] is the size coefficient of the k-th audio channel at the current moment; SUM is the sum of W[k][j] over the range of j, where the value of j is in the range of [0,30 seconds / T]; M[k] is the participation weight of the k-th audio channel at the current moment.

9. The method for multi-person video conference screen layout based on audio analysis according to claim 8, characterized in that: For screen size calculation: The total width of the current screen is WIDTH, and the total height is HEIGHT; For a user with sequence number k, its size coefficient is Q[k] and its ranking coefficient is P[k]; Let the serial number of the current main speaker be m, let the sum of the size coefficients of all users be Qs, and let the number of users whose size coefficients are not 0 except the main speaker be r; The display width width[k] of the screen of user number k; If k == m, then width[k] = (0.565-0.65*r)*WIDTH; If k! =m, then width[k]=(0.435-0.65*r)*WIDTH*Q[k] / (Qs–Q[m]); The display height of the screen of user number k is height[k]; height[k]=(k==m||Q[k]>0)? HEIGHT:

0.

10. The method for multi-person video conference screen layout based on audio analysis according to claim 9, characterized in that: For screen position determination: For users with sequence number k, the ranking coefficient is P[k]; The main speaker screen is placed on the far left; For other users with width[k]>0, they are arranged from right to left according to P[k] from large to small, starting from the right side of the main speaker.

Images