Similar to RoIHeads.postprocess_detections, but can handle embeddings and implement First Classification Score (FCS).

1x1 convolution

Args: use_gt (bool, optional): Whether to use GT as detection results to verify the upper bound of person

Get boxes from proposals.

For each image, load its annotation that is a dictionary with the following keys: img_name (str): image name img_path

Load the image names for the specific split.

Load the image names for the specific split.

Create a small table using the keys of small_dict as headers. This is only suitable for small dictionaries. Args: small_dict (di

gallery_det (list of ndarray): n_det x [x1, y1, x2, y2, score] per image det_thresh (float): filter out gallery detections whose scores below

Get a copy of the default config.

Print dataset statistics.

Reduce the values in the dictionary from all processes so that all processes have the averaged results. Returns a dict with the same fields a

This function disables printing when not in master process

Args: img (Tensor): Tensor image of size (C, H, W). Returns: Tensor: Image with n_holes of dimension length x