Hello, when I replicated your Table 1 experiment with BevFusion using real data, I achieved consistent results with you. However, when we upsampled the 224x400 images to 900x1600 for testing the validation set, the performance AP dropped to 9.1 while your result was 12.3. Could you provide some suggestions?
Hello, when I replicated your Table 1 experiment with BevFusion using real data, I achieved consistent results with you. However, when we upsampled the 224x400 images to 900x1600 for testing the validation set, the performance AP dropped to 9.1 while your result was 12.3. Could you provide some suggestions?