Chengjian Feng

https://drive.google.com/file/d/1dZQ5ytHgJPv4VgYOyjJerq4adc6GQkkd/view?usp=sharing

@jihwanp We conduct the training of regional prompt learning in an off-line manner, and you can find the code [here](https://github.com/fcjian/RPL).

We first get the object features by applying the RoI-align to the FPN features based on the ground-truth bounding boxes. Then, we classify the object features with the classification head...

I had tried to add resume_from_checkpoint or ckpt_path, but it yields the following error: KeyError: 'Trying to restore optimizer state but checkpoint contains only the model. This is probably due...

We also cannot find the both fields of 'camera_synced_box' and 'most_visible_camera_name' in the tfrecord file from waymo data v1.3.1. Are you sure the two fields are included in the new...

+1

@xiaofeng94 How did you set the score thresh to filter the bbox from CLIP and the nms iou thresh to merge the bboxes from CLIP and STAC for the semi-supervised...

The author set the score threshold to 0.9 in STAC paper while you set the score threshold to 0.8 in OVD experiment. I want to know what threshold you set...

@hjfdsssdg Sorry, we have not exported the onnx file.

@Lkevin20 不同层的特征具有不同的感受野大小，从而提取不同尺度的信息。