The prediction was incorrect where use best_model_trancos_ResFCN.pth

[quuhua911]

When I tested it with pictures from the Internet, the results were completely incorrect. What was the reason? python main.py -image_path /home/quh/pythonwork/C-3-Framework/datasets/test/a1/a1.png -model_path checkpoints/best_model_trancos_ResFCN.pth -model_name ResFCN python main.py -image_path /home/quh/pythonwork/C-3-Framework/datasets/test/a1/t1.png -model_path checkpoints/best_model_trancos_ResFCN.pth -model_name ResFCN

[IssamLaradji]

this is expected, because the model is trained on trancos. The dataset only has around 400 images and look very different from the images above.

The model needs to be trained on a dataset like COCO for it to work on internet images like the ones you showed.

Would you like to try training LCFCN on COCO?