ResNet-101 network

[samahwaleed]

1) I want to use resnet-101 model but I can't find it in the models folder. Could you please help me!?

2) How I can find the precision and recall for each object?

Thank you

[sovit-123]

@samahwaleed ResNet101 model is not yet added. Will add that for sure if that is a requirement. Regarding precision and recall. You can check precision and recall, you can visualize class-wise precision and recall using the eval.py script.

[samahwaleed]

I would be grateful if you could add ResNet101 as soon as possible as I need it for research purpose.

Regarding precision and recall. You can check precision and recall, you can visualize class-wise precision and recall using the eval.py script. Could you please explain more how to do it as I need these values for each class!?

Thank you for your time and consideration.

[sovit-123]

Sure, I would add the model by tomorrow.

After training the model, use the eval.py script to check the metrics. Example command: python eval.py --weights outputs/training/custom_training/best_model.pth --config data_configs/custom_data.yaml --model fasterrcnn_resnet50_fpn_v2 --verbose

[samahwaleed]

Thank you so much Sir,

python eval.py --weights outputs/training/custom_training/best_model.pth --config data_configs/custom_data.yaml --model fasterrcnn_resnet50_fpn_v2 --verbose

I have tried this and I only got the AP. How I can get the Recall for each class and F1 score?

[sovit-123]

Right now, the library uses Torchmetrics to show class-wise mAP. I will try to add classwise recall and precision. But it may take some time.

[samahwaleed]

It shows the average precision for each class but not recall, I just need the recall and F1 score.

Thank you so much.

[sovit-123]

I can add recall per class but not sure about the F1 score. Before that, I will add the ResNet101 model.

[sovit-123]

@samahwaleed I have added the ResNet101 model.

[samahwaleed]

Thank you Sir.

I would be thankful if you add the recall at the earliest possible time.