search

cvg / LightGlue

LightGlue: Local Feature Matching at Light Speed (ICCV 2023)
Apache License 2.0
3.15k stars 291 forks source link

I have a problem after training is complete #88

Open qiao89757 opened 7 months ago

qiao89757 commented 7 months ago

Hello, thank you very much for your open source code. I am currently encountering a problem. When I train image registration at a large angle, the configuration file is as follows: data: name: homographies data_dir: revisitop1m train_size: 150000 val_size: 2000 batch_size: 128 num_workers: 14 homography: difficulty: 0.7 max_angle: 180 photometric: name: lg After I trained and fine-tuned, I loaded the generated trained model into lightglue. Even if I ran the code with the same image, the effect of the generated image was as 2023-11-14 10-01-25屏幕截图

Phil26AT commented 5 months ago

Hey, sorry for the late reply. It would be great if you could elaborate your setup with more details. Did the training converge? What is the final loss? How did you load the weights? In your example, both images are identical, so even the pretrained LG should give very good matches.