Open huangchaoxing opened 5 years ago
Hi The paper mentions that in the action recognition dataset , there are 15 subjects for totally 451 videos. However, it seems it only contains 10 subjects for each action. Besides, what is the benchmark split of the dataset ? Thank you
Hi The paper mentions that in the action recognition dataset , there are 15 subjects for totally 451 videos. However, it seems it only contains 10 subjects for each action. Besides, what is the benchmark split of the dataset ? Thank you