Dominik Schörkhuber

TransT tracker integration

2

Hi, I've recently started to use CVAT to annotate videos of traffic scenes. Before coming across CVAT I briefly experimented with Supervisely and was very impressed by their [semi-automatic video...

Random clip per video

12

Hi, I'm currently experimenting with action recognition on the Kinetics 400 dataset. For training an action recognition model I would like to extract a single (sub-) clip per video in...

AI tracker one frame late & TransT integration

5

fixes #28

TransT tracker integration

11

as originally posted in https://github.com/openvinotoolkit/cvat/issues/4768 Hi, I've recently started to use CVAT to annotate videos of traffic scenes. Before coming across CVAT I briefly experimented with Supervisely and was very...

When using the AI tracking tool SiamMask (or any other tracking tool), the tracker is always one frame late. E.g.: 1) I draw a bounding box at frame 0, then...

TransT tracker integration

7

PR to integrate the single object tracker TransT as an AI tool into CVAT. also see here: https://github.com/cvat-ai/cvat-opencv/issues/14

[OcrPlugin] Language Download

2

- [x] Remove language files from this repo - [x] Add action to open language download dialog - [x] Add a list of languages + download buttons - [ ]...

Estimate pose data

1

Hi, you are stating in the Readme that you were not allowed to publish the dataset. However, could you publish the code to estimate the SMPL pose data? Best regards

Hi, I'm experimenting with your code and found that data preprocessing seems to have a pretty large impact on the accuracy. For example [here ](https://github.com/kennymckormick/pyskl/blob/main/tools/data/ntu_preproc.py#L64) it seems you are filtering...