Guo Chen

To extract the RGB and Flow features, you can use this repository [TSN-yjxiong](https://github.com/yjxiong/temporal-segment-networks). And also, you can use [mmaction2](https://mmaction2.readthedocs.io/zh_CN/latest/supported_datasets.html#activitynet) and follow the instruction to extract the TSN feature. Before that,...

@gracikk-ds Hello. We did use InternVL text encoder with 7B parameters for grounding tasks.

@tiesanguaixia Hello, we have released the extracted features at [here](https://huggingface.co/cg1177). You can download them and replace the original features used by CG-DETR with them. You may need to modify some...

@gracikk-ds Could you explain the plot?

@gracikk-ds I believe it is resonable. When I began to train the grounding tasks, stage_2 model was under training. So stage_2_clip 's initialization weight did not have the best video...

> @Andy1621, @cg1177, @LarryLeeee, hi! Any comments about the audio? Hi, I would like to invite another co-author responsible for the audio to answer questions, which will take some time...

> @cg1177, we are limited in time, the conference submission deadline is approaching. Do you have a rough idea of ​​how long it will take to communicate with co-author? We...