The Baidu AI conference was held in Beijing to release new products Apollo and Duer OS
2018-01-30 14:32

In this dataset, Baidu provides more than 1470 full variety shows with their playback links and files containing extracted features. The time stamps of wonderful clips are labeled accurately for each video. The researchers can design algorithms and train models using these videos and labeled clips in the training dataset. The evaluation of models or algorithms is done based on the deviation between the time stamps in the submitted results and the time stamps in our ground truth.

In this dataset, Baidu provides more than 1470 full variety shows with their playback links and files containing extracted features. The time stamps of wonderful clips are labeled accurately for each video. The researchers can design algorithms and train models using these videos and labeled clips in the training dataset. The evaluation of models or algorithms is done based on the deviation between the time stamps in the submitted results and the time stamps in our ground truth.

The Baidu AI conference was held in Beijing to release new products Apollo and Duer OS
2018-01-30 14:33

In this dataset, Baidu provides more than 1470 full variety shows with their playback links and files containing extracted features. The time stamps of wonderful clips are labeled accurately for each video. The researchers can design algorithms and train models using these videos and labeled clips in the training dataset. The evaluation of models or algorithms is done based on the deviation between the time stamps in the submitted results and the time stamps in our ground truth.

In this dataset, Baidu provides more than 1470 full variety shows with their playback links and files containing extracted features. The time stamps of wonderful clips are labeled accurately for each video. The researchers can design algorithms and train models using these videos and labeled clips in the training dataset. The evaluation of models or algorithms is done based on the deviation between the time stamps in the submitted results and the time stamps in our ground truth.