Махлярчук Андрей

Hello @Shanmukh-C Which version of torchvision and pytorch are you using? I notice that for detection demo.py you should use weights of the SSDLite model that is available via [link](https://sc.link/YXg2).

Hello @abhirajasp Maybe you are using a different version of torchvision pkg. We are using torchvision==0.12.0 from [requirements.txt](https://github.com/hukenovs/hagrid/blob/master/requirements.txt). Can you please provide a version of torchvision you are using?

Hello, @narendoraiswamy Can you provide your config you are using? Also [this](https://discuss.pytorch.org/t/runtimeerror-received-0-items-of-ancdata/4999/4) thread may help you.

Hello, @kerem0comert You can use our pre-trained backbone for your task, if you have samples for a new gesture that you want to recognize. Example code for creating a new...

Hello @moonlays We got the following results for subsample dataset: map: 0.679301, map_50: 0.935013, map_75: 0.779909.

Hello @moonlays We trained our model from scratch for 100 epochs and then pick model with the best mAP metric on valid dataset. According to our [paper](https://arxiv.org/pdf/2206.08219.pdf) the detector achieved...