-
I am trying to train the Slowfast network on my custom dataset (where I have a single label for each video). I have prepared the dataset files according to the instructions provided in https://github.…
-
Hi, thanks for sharing this great video recognition framework!
I tried to run the Kinetics/SLOW_4x16_R50.yaml and Kinetics/SLOWFAST_4x16_R50.yaml, which FLOPs are **20** and **27** GFLOPs. However…
-
Hi Team,
Thanks for providing updates to run Demo on AVA and KINETIC dataset.
But with new updates, I am facing a problem to run Demo with [Kinetic pickle file](https://dl.fbaipublicfiles.com/p…
-
### Description
Hi , Thanks for the great work. I have trained the model with my own dataset which has just 15 classes and every 15 classes contain 1 hrs positive clip. Then I started training with …
-
Hi,
In [rename_classnames.sh](https://github.com/open-mmlab/mmaction2/blob/master/tools/data/kinetics/rename_classnames.sh), the `(` and ``)` will be converted to `-`.
However, the [parse_file_l…
-
https://openmmlab.oss-accelerate.aliyuncs.com/mmaction/recognition/slowfast/slowfast_r50_256p_4x16x1_256e_kinetics400_rgb/20200728_022505.log.json#
Only get 30% top-1 acc.
-
Hi, thank you for your great work!
I am trying to understand how multigrid works, especially the subbn module.
The `num_features` of split_bn will be different at each long cycle, to aggregate t…
SuX97 updated
4 years ago
-
Hi author,
This is an impressive study and work! I really like the concept of compositional object interaction and the methodology.
I have a question regarding the frames extraction.
While tryi…
-
Hi, I did not find the results of the experiment. I would like to ask how the UCF 101 dataset performs on slowfast in your experiment? And have you achieve the acc in the original paper?
-
Hi, thanks for sharing the excellent works.
Now I am confused with some questions.
I am trying to reproduce the results of 24.9 mAP as you reported on AVA with R50. The question is I am not sure whi…