Could you please provide details of training the model of appearance (spatial-R34 or spatial RX-50) and motion stream (temporal-R34 or temporal RX-50)?
At first, I tested the pre-trained whole model on our custom dataset. The results are reasonable.
And then, I trained the Interactive model with the pre-trained spatial and temporal models on our custom dataset, the test results are still acceptable.
But when I train the whole model from scratch with our data, the performance is pretty poor.
Could you please provide details of training the model of appearance (spatial-R34 or spatial RX-50) and motion stream (temporal-R34 or temporal RX-50)? At first, I tested the pre-trained whole model on our custom dataset. The results are reasonable.![out-00200](https://user-images.githubusercontent.com/11660168/134310936-90d2c847-70fe-453c-88c4-00e51e4e1973.jpg)
And then, I trained the Interactive model with the pre-trained spatial and temporal models on our custom dataset, the test results are still acceptable.![out-00200](https://user-images.githubusercontent.com/11660168/134310994-b23b5d3a-d403-422d-9763-63f7bfde4403.jpg)
But when I train the whole model from scratch with our data, the performance is pretty poor.![out-00200](https://user-images.githubusercontent.com/11660168/134311028-6e8e5313-4781-4f1e-8eac-8dc40d35cbea.jpg)