# Copyright (c) OpenMMLab. All rights reserved. from typing import Dict, List, Tuple import torch import torch.nn as nn import torch.nn.functional as F from mmcv.cnn import Linear from mmcv.cnn.bricks.transformer import FFN from mmengine.model import BaseModule from mmengine.structures import InstanceData from torch import Tensor from mmdet.registry import MODELS, TASK_UTILS from mmdet.structures import SampleList from mmdet.structures.bbox import bbox_cxcywh_to_xyxy, bbox_xyxy_to_cxcywh from mmdet.utils import (ConfigType, InstanceList, OptInstanceList, OptMultiConfig, reduce_mean) from ..utils import multi_apply @MODELS.register_module() class DETRHead(BaseModule): r"""Head of DETR. DETR:End-to-End Object Detection with Transformers. More details can be found in the `paper `_ . Args: num_classes (int): Number of categories excluding the background. embed_dims (int): The dims of Transformer embedding. num_reg_fcs (int): Number of fully-connected layers used in `FFN`, which is then used for the regression head. Defaults to 2. sync_cls_avg_factor (bool): Whether to sync the `avg_factor` of all ranks. Default to `False`. loss_cls (:obj:`ConfigDict` or dict): Config of the classification loss. Defaults to `CrossEntropyLoss`. loss_bbox (:obj:`ConfigDict` or dict): Config of the regression bbox loss. Defaults to `L1Loss`. loss_iou (:obj:`ConfigDict` or dict): Config of the regression iou loss. Defaults to `GIoULoss`. train_cfg (:obj:`ConfigDict` or dict): Training config of transformer head. test_cfg (:obj:`ConfigDict` or dict): Testing config of transformer head. init_cfg (:obj:`ConfigDict` or dict, optional): the config to control the initialization. Defaults to None. """ _version = 2 def __init__( self, num_classes: int, embed_dims: int = 256, num_reg_fcs: int = 2, sync_cls_avg_factor: bool = False, loss_cls: ConfigType = dict( type='CrossEntropyLoss', bg_cls_weight=0.1, use_sigmoid=False, loss_weight=1.0, class_weight=1.0), loss_bbox: ConfigType = dict(type='L1Loss', loss_weight=5.0), loss_iou: ConfigType = dict(type='GIoULoss', loss_weight=2.0), train_cfg: ConfigType = dict( assigner=dict( type='HungarianAssigner', match_costs=[ dict(type='ClassificationCost', weight=1.), dict(type='BBoxL1Cost', weight=5.0, box_format='xywh'), dict(type='IoUCost', iou_mode='giou', weight=2.0) ])), test_cfg: ConfigType = dict(max_per_img=100), init_cfg: OptMultiConfig = None) -> None: super().__init__(init_cfg=init_cfg) self.bg_cls_weight = 0 self.sync_cls_avg_factor = sync_cls_avg_factor class_weight = loss_cls.get('class_weight', None) if class_weight is not None and (self.__class__ is DETRHead): assert isinstance(class_weight, float), 'Expected ' \ 'class_weight to have type float. Found ' \ f'{type(class_weight)}.' # NOTE following the official DETR repo, bg_cls_weight means # relative classification weight of the no-object class. bg_cls_weight = loss_cls.get('bg_cls_weight', class_weight) assert isinstance(bg_cls_weight, float), 'Expected ' \ 'bg_cls_weight to have type float. Found ' \ f'{type(bg_cls_weight)}.' class_weight = torch.ones(num_classes + 1) * class_weight # set background class as the last indice class_weight[num_classes] = bg_cls_weight loss_cls.update({'class_weight': class_weight}) if 'bg_cls_weight' in loss_cls: loss_cls.pop('bg_cls_weight') self.bg_cls_weight = bg_cls_weight if train_cfg: assert 'assigner' in train_cfg, 'assigner should be provided ' \ 'when train_cfg is set.' assigner = train_cfg['assigner'] self.assigner = TASK_UTILS.build(assigner) if train_cfg.get('sampler', None) is not None: raise RuntimeError('DETR do not build sampler.') self.num_classes = num_classes self.embed_dims = embed_dims self.num_reg_fcs = num_reg_fcs self.train_cfg = train_cfg self.test_cfg = test_cfg self.loss_cls = MODELS.build(loss_cls) self.loss_bbox = MODELS.build(loss_bbox) self.loss_iou = MODELS.build(loss_iou) if self.loss_cls.use_sigmoid: self.cls_out_channels = num_classes else: self.cls_out_channels = num_classes + 1 self._init_layers() def _init_layers(self) -> None: """Initialize layers of the transformer head.""" # cls branch self.fc_cls = Linear(self.embed_dims, self.cls_out_channels) # reg branch self.activate = nn.ReLU() self.reg_ffn = FFN( self.embed_dims, self.embed_dims, self.num_reg_fcs, dict(type='ReLU', inplace=True), dropout=0.0, add_residual=False) # NOTE the activations of reg_branch here is the same as # those in transformer, but they are actually different # in DAB-DETR (prelu in transformer and relu in reg_branch) self.fc_reg = Linear(self.embed_dims, 4) def forward(self, hidden_states: Tensor) -> Tuple[Tensor]: """"Forward function. Args: hidden_states (Tensor): Features from transformer decoder. If `return_intermediate_dec` in detr.py is True output has shape (num_decoder_layers, bs, num_queries, dim), else has shape (1, bs, num_queries, dim) which only contains the last layer outputs. Returns: tuple[Tensor]: results of head containing the following tensor. - layers_cls_scores (Tensor): Outputs from the classification head, shape (num_decoder_layers, bs, num_queries, cls_out_channels). Note cls_out_channels should include background. - layers_bbox_preds (Tensor): Sigmoid outputs from the regression head with normalized coordinate format (cx, cy, w, h), has shape (num_decoder_layers, bs, num_queries, 4). """ layers_cls_scores = self.fc_cls(hidden_states) layers_bbox_preds = self.fc_reg( self.activate(self.reg_ffn(hidden_states))).sigmoid() return layers_cls_scores, layers_bbox_preds def loss(self, hidden_states: Tensor, batch_data_samples: SampleList) -> dict: """Perform forward propagation and loss calculation of the detection head on the features of the upstream network. Args: hidden_states (Tensor): Feature from the transformer decoder, has shape (num_decoder_layers, bs, num_queries, cls_out_channels) or (num_decoder_layers, num_queries, bs, cls_out_channels). batch_data_samples (List[:obj:`DetDataSample`]): The Data Samples. It usually includes information such as `gt_instance`, `gt_panoptic_seg` and `gt_sem_seg`. Returns: dict: A dictionary of loss components. """ batch_gt_instances = [] batch_img_metas = [] for data_sample in batch_data_samples: batch_img_metas.append(data_sample.metainfo) batch_gt_instances.append(data_sample.gt_instances) outs = self(hidden_states) loss_inputs = outs + (batch_gt_instances, batch_img_metas) losses = self.loss_by_feat(*loss_inputs) return losses def loss_by_feat( self, all_layers_cls_scores: Tensor, all_layers_bbox_preds: Tensor, batch_gt_instances: InstanceList, batch_img_metas: List[dict], batch_gt_instances_ignore: OptInstanceList = None ) -> Dict[str, Tensor]: """"Loss function. Only outputs from the last feature level are used for computing losses by default. Args: all_layers_cls_scores (Tensor): Classification outputs of each decoder layers. Each is a 4D-tensor, has shape (num_decoder_layers, bs, num_queries, cls_out_channels). all_layers_bbox_preds (Tensor): Sigmoid regression outputs of each decoder layers. Each is a 4D-tensor with normalized coordinate format (cx, cy, w, h) and shape (num_decoder_layers, bs, num_queries, 4). batch_gt_instances (list[:obj:`InstanceData`]): Batch of gt_instance. It usually includes ``bboxes`` and ``labels`` attributes. batch_img_metas (list[dict]): Meta information of each image, e.g., image size, scaling factor, etc. batch_gt_instances_ignore (list[:obj:`InstanceData`], optional): Batch of gt_instances_ignore. It includes ``bboxes`` attribute data that is ignored during training and testing. Defaults to None. Returns: dict[str, Tensor]: A dictionary of loss components. """ assert batch_gt_instances_ignore is None, \ f'{self.__class__.__name__} only supports ' \ 'for batch_gt_instances_ignore setting to None.' losses_cls, losses_bbox, losses_iou = multi_apply( self.loss_by_feat_single, all_layers_cls_scores, all_layers_bbox_preds, batch_gt_instances=batch_gt_instances, batch_img_metas=batch_img_metas) loss_dict = dict() # loss from the last decoder layer loss_dict['loss_cls'] = losses_cls[-1] loss_dict['loss_bbox'] = losses_bbox[-1] loss_dict['loss_iou'] = losses_iou[-1] # loss from other decoder layers num_dec_layer = 0 for loss_cls_i, loss_bbox_i, loss_iou_i in \ zip(losses_cls[:-1], losses_bbox[:-1], losses_iou[:-1]): loss_dict[f'd{num_dec_layer}.loss_cls'] = loss_cls_i loss_dict[f'd{num_dec_layer}.loss_bbox'] = loss_bbox_i loss_dict[f'd{num_dec_layer}.loss_iou'] = loss_iou_i num_dec_layer += 1 return loss_dict def loss_by_feat_single(self, cls_scores: Tensor, bbox_preds: Tensor, batch_gt_instances: InstanceList, batch_img_metas: List[dict]) -> Tuple[Tensor]: """Loss function for outputs from a single decoder layer of a single feature level. Args: cls_scores (Tensor): Box score logits from a single decoder layer for all images, has shape (bs, num_queries, cls_out_channels). bbox_preds (Tensor): Sigmoid outputs from a single decoder layer for all images, with normalized coordinate (cx, cy, w, h) and shape (bs, num_queries, 4). batch_gt_instances (list[:obj:`InstanceData`]): Batch of gt_instance. It usually includes ``bboxes`` and ``labels`` attributes. batch_img_metas (list[dict]): Meta information of each image, e.g., image size, scaling factor, etc. Returns: Tuple[Tensor]: A tuple including `loss_cls`, `loss_box` and `loss_iou`. """ num_imgs = cls_scores.size(0) cls_scores_list = [cls_scores[i] for i in range(num_imgs)] bbox_preds_list = [bbox_preds[i] for i in range(num_imgs)] cls_reg_targets = self.get_targets(cls_scores_list, bbox_preds_list, batch_gt_instances, batch_img_metas) (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list, num_total_pos, num_total_neg) = cls_reg_targets labels = torch.cat(labels_list, 0) label_weights = torch.cat(label_weights_list, 0) bbox_targets = torch.cat(bbox_targets_list, 0) bbox_weights = torch.cat(bbox_weights_list, 0) # classification loss cls_scores = cls_scores.reshape(-1, self.cls_out_channels) # construct weighted avg_factor to match with the official DETR repo cls_avg_factor = num_total_pos * 1.0 + \ num_total_neg * self.bg_cls_weight if self.sync_cls_avg_factor: cls_avg_factor = reduce_mean( cls_scores.new_tensor([cls_avg_factor])) cls_avg_factor = max(cls_avg_factor, 1) loss_cls = self.loss_cls( cls_scores, labels, label_weights, avg_factor=cls_avg_factor) # Compute the average number of gt boxes across all gpus, for # normalization purposes num_total_pos = loss_cls.new_tensor([num_total_pos]) num_total_pos = torch.clamp(reduce_mean(num_total_pos), min=1).item() # construct factors used for rescale bboxes factors = [] for img_meta, bbox_pred in zip(batch_img_metas, bbox_preds): img_h, img_w, = img_meta['img_shape'] factor = bbox_pred.new_tensor([img_w, img_h, img_w, img_h]).unsqueeze(0).repeat( bbox_pred.size(0), 1) factors.append(factor) factors = torch.cat(factors, 0) # DETR regress the relative position of boxes (cxcywh) in the image, # thus the learning target is normalized by the image size. So here # we need to re-scale them for calculating IoU loss bbox_preds = bbox_preds.reshape(-1, 4) bboxes = bbox_cxcywh_to_xyxy(bbox_preds) * factors bboxes_gt = bbox_cxcywh_to_xyxy(bbox_targets) * factors # regression IoU loss, defaultly GIoU loss loss_iou = self.loss_iou( bboxes, bboxes_gt, bbox_weights, avg_factor=num_total_pos) # regression L1 loss loss_bbox = self.loss_bbox( bbox_preds, bbox_targets, bbox_weights, avg_factor=num_total_pos) return loss_cls, loss_bbox, loss_iou def get_targets(self, cls_scores_list: List[Tensor], bbox_preds_list: List[Tensor], batch_gt_instances: InstanceList, batch_img_metas: List[dict]) -> tuple: """Compute regression and classification targets for a batch image. Outputs from a single decoder layer of a single feature level are used. Args: cls_scores_list (list[Tensor]): Box score logits from a single decoder layer for each image, has shape [num_queries, cls_out_channels]. bbox_preds_list (list[Tensor]): Sigmoid outputs from a single decoder layer for each image, with normalized coordinate (cx, cy, w, h) and shape [num_queries, 4]. batch_gt_instances (list[:obj:`InstanceData`]): Batch of gt_instance. It usually includes ``bboxes`` and ``labels`` attributes. batch_img_metas (list[dict]): Meta information of each image, e.g., image size, scaling factor, etc. Returns: tuple: a tuple containing the following targets. - labels_list (list[Tensor]): Labels for all images. - label_weights_list (list[Tensor]): Label weights for all images. - bbox_targets_list (list[Tensor]): BBox targets for all images. - bbox_weights_list (list[Tensor]): BBox weights for all images. - num_total_pos (int): Number of positive samples in all images. - num_total_neg (int): Number of negative samples in all images. """ (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list, pos_inds_list, neg_inds_list) = multi_apply(self._get_targets_single, cls_scores_list, bbox_preds_list, batch_gt_instances, batch_img_metas) num_total_pos = sum((inds.numel() for inds in pos_inds_list)) num_total_neg = sum((inds.numel() for inds in neg_inds_list)) return (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list, num_total_pos, num_total_neg) def _get_targets_single(self, cls_score: Tensor, bbox_pred: Tensor, gt_instances: InstanceData, img_meta: dict) -> tuple: """Compute regression and classification targets for one image. Outputs from a single decoder layer of a single feature level are used. Args: cls_score (Tensor): Box score logits from a single decoder layer for one image. Shape [num_queries, cls_out_channels]. bbox_pred (Tensor): Sigmoid outputs from a single decoder layer for one image, with normalized coordinate (cx, cy, w, h) and shape [num_queries, 4]. gt_instances (:obj:`InstanceData`): Ground truth of instance annotations. It should includes ``bboxes`` and ``labels`` attributes. img_meta (dict): Meta information for one image. Returns: tuple[Tensor]: a tuple containing the following for one image. - labels (Tensor): Labels of each image. - label_weights (Tensor]): Label weights of each image. - bbox_targets (Tensor): BBox targets of each image. - bbox_weights (Tensor): BBox weights of each image. - pos_inds (Tensor): Sampled positive indices for each image. - neg_inds (Tensor): Sampled negative indices for each image. """ img_h, img_w = img_meta['img_shape'] factor = bbox_pred.new_tensor([img_w, img_h, img_w, img_h]).unsqueeze(0) num_bboxes = bbox_pred.size(0) # convert bbox_pred from xywh, normalized to xyxy, unnormalized bbox_pred = bbox_cxcywh_to_xyxy(bbox_pred) bbox_pred = bbox_pred * factor pred_instances = InstanceData(scores=cls_score, bboxes=bbox_pred) # assigner and sampler assign_result = self.assigner.assign( pred_instances=pred_instances, gt_instances=gt_instances, img_meta=img_meta) gt_bboxes = gt_instances.bboxes gt_labels = gt_instances.labels pos_inds = torch.nonzero( assign_result.gt_inds > 0, as_tuple=False).squeeze(-1).unique() neg_inds = torch.nonzero( assign_result.gt_inds == 0, as_tuple=False).squeeze(-1).unique() pos_assigned_gt_inds = assign_result.gt_inds[pos_inds] - 1 pos_gt_bboxes = gt_bboxes[pos_assigned_gt_inds.long(), :] # label targets labels = gt_bboxes.new_full((num_bboxes, ), self.num_classes, dtype=torch.long) labels[pos_inds] = gt_labels[pos_assigned_gt_inds] label_weights = gt_bboxes.new_ones(num_bboxes) # bbox targets bbox_targets = torch.zeros_like(bbox_pred) bbox_weights = torch.zeros_like(bbox_pred) bbox_weights[pos_inds] = 1.0 # DETR regress the relative position of boxes (cxcywh) in the image. # Thus the learning target should be normalized by the image size, also # the box format should be converted from defaultly x1y1x2y2 to cxcywh. pos_gt_bboxes_normalized = pos_gt_bboxes / factor pos_gt_bboxes_targets = bbox_xyxy_to_cxcywh(pos_gt_bboxes_normalized) bbox_targets[pos_inds] = pos_gt_bboxes_targets return (labels, label_weights, bbox_targets, bbox_weights, pos_inds, neg_inds) def loss_and_predict( self, hidden_states: Tuple[Tensor], batch_data_samples: SampleList) -> Tuple[dict, InstanceList]: """Perform forward propagation of the head, then calculate loss and predictions from the features and data samples. Over-write because img_metas are needed as inputs for bbox_head. Args: hidden_states (tuple[Tensor]): Feature from the transformer decoder, has shape (num_decoder_layers, bs, num_queries, dim). batch_data_samples (list[:obj:`DetDataSample`]): Each item contains the meta information of each image and corresponding annotations. Returns: tuple: the return value is a tuple contains: - losses: (dict[str, Tensor]): A dictionary of loss components. - predictions (list[:obj:`InstanceData`]): Detection results of each image after the post process. """ batch_gt_instances = [] batch_img_metas = [] for data_sample in batch_data_samples: batch_img_metas.append(data_sample.metainfo) batch_gt_instances.append(data_sample.gt_instances) outs = self(hidden_states) loss_inputs = outs + (batch_gt_instances, batch_img_metas) losses = self.loss_by_feat(*loss_inputs) predictions = self.predict_by_feat( *outs, batch_img_metas=batch_img_metas) return losses, predictions def predict(self, hidden_states: Tuple[Tensor], batch_data_samples: SampleList, rescale: bool = True) -> InstanceList: """Perform forward propagation of the detection head and predict detection results on the features of the upstream network. Over-write because img_metas are needed as inputs for bbox_head. Args: hidden_states (tuple[Tensor]): Multi-level features from the upstream network, each is a 4D-tensor. batch_data_samples (List[:obj:`DetDataSample`]): The Data Samples. It usually includes information such as `gt_instance`, `gt_panoptic_seg` and `gt_sem_seg`. rescale (bool, optional): Whether to rescale the results. Defaults to True. Returns: list[obj:`InstanceData`]: Detection results of each image after the post process. """ batch_img_metas = [ data_samples.metainfo for data_samples in batch_data_samples ] last_layer_hidden_state = hidden_states[-1].unsqueeze(0) outs = self(last_layer_hidden_state) predictions = self.predict_by_feat( *outs, batch_img_metas=batch_img_metas, rescale=rescale) return predictions def predict_by_feat(self, layer_cls_scores: Tensor, layer_bbox_preds: Tensor, batch_img_metas: List[dict], rescale: bool = True) -> InstanceList: """Transform network outputs for a batch into bbox predictions. Args: layer_cls_scores (Tensor): Classification outputs of the last or all decoder layer. Each is a 4D-tensor, has shape (num_decoder_layers, bs, num_queries, cls_out_channels). layer_bbox_preds (Tensor): Sigmoid regression outputs of the last or all decoder layer. Each is a 4D-tensor with normalized coordinate format (cx, cy, w, h) and shape (num_decoder_layers, bs, num_queries, 4). batch_img_metas (list[dict]): Meta information of each image. rescale (bool, optional): If `True`, return boxes in original image space. Defaults to `True`. Returns: list[:obj:`InstanceData`]: Object detection results of each image after the post process. Each item usually contains following keys. - scores (Tensor): Classification scores, has a shape (num_instance, ) - labels (Tensor): Labels of bboxes, has a shape (num_instances, ). - bboxes (Tensor): Has a shape (num_instances, 4), the last dimension 4 arrange as (x1, y1, x2, y2). """ # NOTE only using outputs from the last feature level, # and only the outputs from the last decoder layer is used. cls_scores = layer_cls_scores[-1] bbox_preds = layer_bbox_preds[-1] result_list = [] for img_id in range(len(batch_img_metas)): cls_score = cls_scores[img_id] bbox_pred = bbox_preds[img_id] img_meta = batch_img_metas[img_id] results = self._predict_by_feat_single(cls_score, bbox_pred, img_meta, rescale) result_list.append(results) return result_list def _predict_by_feat_single(self, cls_score: Tensor, bbox_pred: Tensor, img_meta: dict, rescale: bool = True) -> InstanceData: """Transform outputs from the last decoder layer into bbox predictions for each image. Args: cls_score (Tensor): Box score logits from the last decoder layer for each image. Shape [num_queries, cls_out_channels]. bbox_pred (Tensor): Sigmoid outputs from the last decoder layer for each image, with coordinate format (cx, cy, w, h) and shape [num_queries, 4]. img_meta (dict): Image meta info. rescale (bool): If True, return boxes in original image space. Default True. Returns: :obj:`InstanceData`: Detection results of each image after the post process. Each item usually contains following keys. - scores (Tensor): Classification scores, has a shape (num_instance, ) - labels (Tensor): Labels of bboxes, has a shape (num_instances, ). - bboxes (Tensor): Has a shape (num_instances, 4), the last dimension 4 arrange as (x1, y1, x2, y2). """ assert len(cls_score) == len(bbox_pred) # num_queries max_per_img = self.test_cfg.get('max_per_img', len(cls_score)) img_shape = img_meta['img_shape'] # exclude background if self.loss_cls.use_sigmoid: cls_score = cls_score.sigmoid() scores, indexes = cls_score.view(-1).topk(max_per_img) det_labels = indexes % self.num_classes bbox_index = indexes // self.num_classes bbox_pred = bbox_pred[bbox_index] else: scores, det_labels = F.softmax(cls_score, dim=-1)[..., :-1].max(-1) scores, bbox_index = scores.topk(max_per_img) bbox_pred = bbox_pred[bbox_index] det_labels = det_labels[bbox_index] det_bboxes = bbox_cxcywh_to_xyxy(bbox_pred) det_bboxes[:, 0::2] = det_bboxes[:, 0::2] * img_shape[1] det_bboxes[:, 1::2] = det_bboxes[:, 1::2] * img_shape[0] det_bboxes[:, 0::2].clamp_(min=0, max=img_shape[1]) det_bboxes[:, 1::2].clamp_(min=0, max=img_shape[0]) if rescale: assert img_meta.get('scale_factor') is not None det_bboxes /= det_bboxes.new_tensor( img_meta['scale_factor']).repeat((1, 2)) results = InstanceData() results.bboxes = det_bboxes results.scores = scores results.labels = det_labels return results