用自己图片运行grounded_sam_demo.py

[zqwang21]

为什么只识别到一个车辆，其他车辆没有识别到呢？

运行命令： export CUDA_VISIBLE_DEVICES=0 python grounded_sam_demo.py \ --config GroundingDINO/groundingdino/config/GroundingDINO_SwinT_OGC.py \ --grounded_checkpoint groundingdino_swint_ogc.pth \ --sam_checkpoint sam_vit_h_4b8939.pth \ --input_image assets/P00111024343___0.png \ --output_dir "outputs" \ --box_threshold 0.3 \ --text_threshold 0.25 \ --text_prompt "large-vehicle.small-vehicle." \ --device "cuda"

prompt换成car和truck之后 效果也不理想 也有很多large-vehicle没有检测出来（只把货车头检测为car 但是车厢没有检测出来），请问有什么好的解决办法吗？

[rentainhe]

Our model has not seen many images in such scenarios at the moment, and we will continue to enhance the model's capabilities to address this situation.