Determining tactical relevance and similarity of video sequences
Methods and systems for finding and ranking video sequences that contain tactical sequences (e.g., an uninterrupted portion of a sporting event) that are similar to tactical event content in a target video sequence, such as those occurring in sports, include learning and labelling (e.g., through use of metadata and the like) tactically significant sequences in a plurality of video segments, including a plurality of such tactically significant sequences in a single video segment.
1. A computer-implemented method for determining tactical similarity of video sequences depicting sporting activity, the method comprising:
receiving, at one or more processors, a first video feed, which depicts a sporting event;
analysing, with the one or more processors, spatiotemporal data corresponding to the sporting event;
identifying, with the one or more processors, a first video sequence within the video feed, based on the analysing of the spatiotemporal data;
generating, with the one or more processors, one or more semantic labels for the first video sequence, based on the analysing of the spatiotemporal data; and
identifying, with the one or more processors, from a library of stored video sequences, one or more stored video sequences that are tactically similar to the first video sequence, using a tactical similarity detection algorithm and the one or more semantic labels for the first video sequence.
2. The method of claim 1 , further comprising determining, with the one or more processors, based on the spatiotemporal data, a location at which a tactical event depicted in the first video sequence occurs,
wherein the tactical similarity detection algorithm compares the location of the tactical event depicted in the first video sequence with locations at which tactical events depicted in the stored video sequences occurred.
3. The method of claim 1 , wherein said analysing of the spatiotemporal data utilises a machine learning algorithm.
4. The method of claim 1 , wherein the tactical similarity detection algorithm utilises a reference data set representative of tactical similarity factors.
5. The method of claim 1 , further comprising determining, with the one or more processors, tactical weighting factors for one or more elements depicted in the first video sequence,
wherein the tactical similarity detection algorithm utilises the tactical weighting factors for the one or more elements depicted in the first video sequence and tactical weighting factors for elements depicted in the stored video sequences.
6. The method of claim 5 , wherein at least some of said one or more elements depicted in the first video sequence are players in the sporting event, and at least some of said elements depicted in the stored video sequences are players in other sporting events.
7. The method of claim 1 , further comprising determining, with the one or more processors, a degree of tactical effect of a tactical event depicted in the first video sequence,
wherein the tactical similarity detection algorithm compares the degree of tactical effect of the tactical event depicted in the first video sequence with degrees of tactical effect of tactical events depicted in the stored video sequences.
8. A system for determining tactical similarity of video sequences depicting sporting activity, the system comprising:
one or more computer processors;
memory storing computer-executable instructions that, when executed by the one or more computer processors, cause the one or more computer processors to perform operations comprising:
receiving a first video feed, which depicts a sporting event;
analysing spatiotemporal data corresponding to the sporting event;
detecting a first video sequence within the video feed, based on the analysing of the spatiotemporal data;
generating one or more semantic labels for the first video sequence, based on the analysing of the spatiotemporal data; and
identifying from a library of stored video sequences one or more stored video sequences that are tactically similar to the first video sequence, using a tactical similarity detection algorithm and the one or more semantic labels for the first video sequence.
9. The system of claim 8 , wherein the operations further comprise determining, based on the spatiotemporal data, a location at which a tactical event depicted in the first video sequence occurs,
wherein the tactical similarity detection algorithm compares the location of the tactical event depicted in the first video sequence with locations at which tactical events depicted in the stored video sequences occurred.
10. The system of claim 8 , wherein said analysing of the spatiotemporal data utilises a machine learning algorithm.
11. The system of claim 8 , wherein the tactical similarity detection algorithm utilises a reference data set representative of tactical similarity factors.
12. The system of claim 8 , wherein the operations further comprise determining tactical weighting factors for one or more elements depicted in the first video sequence,
wherein the tactical similarity detection algorithm utilises the tactical weighting factors for the one or more elements depicted in the first video sequence and tactical weighting factors for elements depicted in the stored video sequences.
13. The system of claim 8 , wherein the operations further comprise determining a degree of tactical effect of a tactical event depicted in the first video sequence,
wherein the tactical similarity detection algorithm compares the degree of tactical effect of the tactical event depicted in the first video sequence with degrees of tactical effect of tactical events depicted in the stored video sequences.
14. A non-transitory computer-readable medium storing instructions that adapt at least one processor to carry out operations comprising:
receiving a first video feed, which depicts a sporting event;
analysing spatiotemporal data corresponding to the sporting event;
detecting a first video sequence within the video feed, based on the analysing of the spatiotemporal data;
generating one or more semantic labels for the first video sequence, based on the analysing of the spatiotemporal data; and
identifying from a library of stored video sequences one or more stored video sequences that are tactically similar to the first video sequence, using a tactical similarity detection algorithm and the one or more semantic labels for the first video sequence.
15. The computer-readable medium of claim 14 , wherein the operations further comprise determining, based on the spatiotemporal data, a location at which a tactical event depicted in the first video sequence occurs,
wherein the tactical similarity detection algorithm compares the location of the tactical event depicted in the first video sequence with locations at which tactical events depicted in the stored video sequences occurred.
16. The computer-readable medium of claim 14 , wherein said analysing of the spatiotemporal data utilises a machine learning algorithm.
17. The computer-readable medium of claim 14 , wherein the tactical similarity detection algorithm utilises a reference data set representative of tactical similarity factors.
18. The computer-readable medium of claim 14 , wherein the operations further comprise determining tactical weighting factors for one or more elements depicted in the first video sequence,
wherein the tactical similarity detection algorithm utilises the tactical weighting factors for the one or more elements depicted in the first video sequence and tactical weighting factors for elements depicted in the stored video sequences.
19. The computer-readable medium of claim 18 , wherein at least some of said one or more elements depicted in the first video sequence are players in the sporting event, and at least some of said elements depicted in the stored video sequences are players in other sporting events.
20. The computer-readable medium of claim 14 , wherein the operations further comprise determining a degree of tactical effect of a tactical event depicted in the first video sequence,
wherein the tactical similarity detection algorithm compares the degree of tactical effect of the tactical event depicted in the first video sequence with degrees of tactical effect of tactical events depicted in the stored video sequences.