# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. from __future__ import absolute_import from __future__ import division from __future__ import print_function from collections import OrderedDict import copy import numpy as np import paddle.fluid as fluid from paddle.fluid.param_attr import ParamAttr from paddle.fluid.initializer import MSRA from paddle.fluid.regularizer import L2Decay from ppdet.experimental import mixed_precision_global_state from ppdet.core.workspace import register from .input_helper import multiscale_def __all__ = ['HybridTaskCascade'] @register class HybridTaskCascade(object): """ Hybrid Task Cascade Mask R-CNN architecture, see https://arxiv.org/abs/1901.07518 Args: backbone (object): backbone instance rpn_head (object): `RPNhead` instance bbox_assigner (object): `BBoxAssigner` instance roi_extractor (object): ROI extractor instance bbox_head (object): `HTCBBoxHead` instance mask_assigner (object): `MaskAssigner` instance mask_head (object): `HTCMaskHead` instance fpn (object): feature pyramid network instance semantic_roi_extractor(object): ROI extractor instance fused_semantic_head (object): `FusedSemanticHead` instance """ __category__ = 'architecture' __inject__ = [ 'backbone', 'rpn_head', 'bbox_assigner', 'roi_extractor', 'bbox_head', 'mask_assigner', 'mask_head', 'fpn', 'semantic_roi_extractor', 'fused_semantic_head' ] def __init__(self, backbone, rpn_head, roi_extractor='FPNRoIAlign', semantic_roi_extractor='RoIAlign', fused_semantic_head='FusedSemanticHead', bbox_head='HTCBBoxHead', bbox_assigner='CascadeBBoxAssigner', mask_assigner='MaskAssigner', mask_head='HTCMaskHead', rpn_only=False, fpn='FPN'): super(HybridTaskCascade, self).__init__() assert fpn is not None, "HTC requires FPN" self.backbone = backbone self.fpn = fpn self.rpn_head = rpn_head self.bbox_assigner = bbox_assigner self.roi_extractor = roi_extractor self.semantic_roi_extractor = semantic_roi_extractor self.fused_semantic_head = fused_semantic_head self.bbox_head = bbox_head self.mask_assigner = mask_assigner self.mask_head = mask_head self.rpn_only = rpn_only # Cascade local cfg self.cls_agnostic_bbox_reg = 2 (brw0, brw1, brw2) = self.bbox_assigner.bbox_reg_weights self.cascade_bbox_reg_weights = [ [1. / brw0, 1. / brw0, 2. / brw0, 2. / brw0], [1. / brw1, 1. / brw1, 2. / brw1, 2. / brw1], [1. / brw2, 1. / brw2, 2. / brw2, 2. / brw2] ] self.cascade_rcnn_loss_weight = [1.0, 0.5, 0.25] self.num_stage = 3 self.with_mask = True self.interleaved = True self.mask_info_flow = True self.with_semantic = True self.use_bias_scalar = True def build(self, feed_vars, mode='train'): if mode == 'train': required_fields = [ 'gt_class', 'gt_bbox', 'gt_mask', 'is_crowd', 'im_info', 'semantic' ] else: required_fields = ['im_shape', 'im_info'] self._input_check(required_fields, feed_vars) im = feed_vars['image'] if mode == 'train': gt_bbox = feed_vars['gt_bbox'] is_crowd = feed_vars['is_crowd'] im_info = feed_vars['im_info'] # backbone body_feats = self.backbone(im) loss = {} # FPN if self.fpn is not None: body_feats, spatial_scale = self.fpn.get_output(body_feats) if self.with_semantic: # TODO: use cfg semantic_feat, seg_pred = self.fused_semantic_head.get_out( body_feats) if mode == 'train': s_label = feed_vars['semantic'] semantic_loss = self.fused_semantic_head.get_loss(seg_pred, s_label) * 0.2 loss.update({"semantic_loss": semantic_loss}) else: semantic_feat = None # rpn proposals rpn_rois = self.rpn_head.get_proposals(body_feats, im_info, mode=mode) if mode == 'train': rpn_loss = self.rpn_head.get_loss(im_info, gt_bbox, is_crowd) loss.update(rpn_loss) else: if self.rpn_only: im_scale = fluid.layers.slice( im_info, [1], starts=[2], ends=[3]) im_scale = fluid.layers.sequence_expand(im_scale, rpn_rois) rois = rpn_rois / im_scale return {'proposal': rois} proposal_list = [] roi_feat_list = [] rcnn_pred_list = [] rcnn_target_list = [] mask_logits_list = [] mask_target_list = [] proposals = None bbox_pred = None outs = None refined_bbox = rpn_rois max_overlap = None for i in range(self.num_stage): # BBox Branch if mode == 'train': outs = self.bbox_assigner( input_rois=refined_bbox, feed_vars=feed_vars, curr_stage=i, max_overlap=max_overlap) proposals = outs[0] max_overlap = outs[-1] rcnn_target_list.append(outs[:-1]) else: proposals = refined_bbox proposal_list.append(proposals) # extract roi features roi_feat = self.roi_extractor(body_feats, proposals, spatial_scale) if self.with_semantic: semantic_roi_feat = self.semantic_roi_extractor(semantic_feat, proposals) if semantic_roi_feat is not None: semantic_roi_feat = fluid.layers.pool2d( semantic_roi_feat, pool_size=2, pool_stride=2, pool_padding='SAME') roi_feat = fluid.layers.sum([roi_feat, semantic_roi_feat]) roi_feat_list.append(roi_feat) # bbox head cls_score, bbox_pred = self.bbox_head.get_output( roi_feat, wb_scalar=1.0 / self.cascade_rcnn_loss_weight[i], name='_' + str(i)) rcnn_pred_list.append((cls_score, bbox_pred)) # Mask Branch if self.with_mask: if mode == 'train': labels_int32 = outs[1] if self.interleaved: refined_bbox = self._decode_box( proposals, bbox_pred, curr_stage=i) proposals = refined_bbox mask_rois, roi_has_mask_int32, mask_int32 = self.mask_assigner( rois=proposals, gt_classes=feed_vars['gt_class'], is_crowd=feed_vars['is_crowd'], gt_segms=feed_vars['gt_mask'], im_info=feed_vars['im_info'], labels_int32=labels_int32) mask_target_list.append(mask_int32) mask_feat = self.roi_extractor( body_feats, mask_rois, spatial_scale, is_mask=True) if self.with_semantic: semantic_roi_feat = self.semantic_roi_extractor( semantic_feat, mask_rois) if semantic_roi_feat is not None: mask_feat = fluid.layers.sum( [mask_feat, semantic_roi_feat]) if self.mask_info_flow: last_feat = None for j in range(i): last_feat = self.mask_head.get_output( mask_feat, last_feat, return_logits=False, return_feat=True, wb_scalar=1.0 / self.cascade_rcnn_loss_weight[i] if self.use_bias_scalar else 1.0, name='_' + str(i) + '_' + str(j)) mask_logits = self.mask_head.get_output( mask_feat, last_feat, return_logits=True, return_feat=False, wb_scalar=1.0 / self.cascade_rcnn_loss_weight[i] if self.use_bias_scalar else 1.0, name='_' + str(i)) else: mask_logits = self.mask_head.get_output( mask_feat, return_logits=True, wb_scalar=1.0 / self.cascade_rcnn_loss_weight[i] if self.use_bias_scalar else 1.0, name='_' + str(i)) mask_logits_list.append(mask_logits) if i < self.num_stage - 1 and not self.interleaved: refined_bbox = self._decode_box( proposals, bbox_pred, curr_stage=i) elif i < self.num_stage - 1 and mode != 'train': refined_bbox = self._decode_box( proposals, bbox_pred, curr_stage=i) if mode == 'train': bbox_loss = self.bbox_head.get_loss( rcnn_pred_list, rcnn_target_list, self.cascade_rcnn_loss_weight) loss.update(bbox_loss) mask_loss = self.mask_head.get_loss(mask_logits_list, mask_target_list, self.cascade_rcnn_loss_weight) loss.update(mask_loss) total_loss = fluid.layers.sum(list(loss.values())) loss.update({'loss': total_loss}) return loss else: mask_name = 'mask_pred' mask_pred, bbox_pred = self.single_scale_eval( body_feats, spatial_scale, im_info, mask_name, bbox_pred, roi_feat_list, rcnn_pred_list, proposal_list, feed_vars['im_shape'], semantic_feat=semantic_feat if self.with_semantic else None) return {'bbox': bbox_pred, 'mask': mask_pred} def single_scale_eval(self, body_feats, spatial_scale, im_info, mask_name, bbox_pred, roi_feat_list=None, rcnn_pred_list=None, proposal_list=None, im_shape=None, use_multi_test=False, semantic_feat=None): if not use_multi_test: bbox_pred = self.bbox_head.get_prediction( im_info, im_shape, roi_feat_list, rcnn_pred_list, proposal_list, self.cascade_bbox_reg_weights) bbox_pred = bbox_pred['bbox'] # share weight bbox_shape = fluid.layers.shape(bbox_pred) bbox_size = fluid.layers.reduce_prod(bbox_shape) bbox_size = fluid.layers.reshape(bbox_size, [1, 1]) size = fluid.layers.fill_constant([1, 1], value=6, dtype='int32') cond = fluid.layers.less_than(x=bbox_size, y=size) mask_pred = fluid.layers.create_global_var( shape=[1], value=0.0, dtype='float32', persistable=False, name=mask_name) def noop(): fluid.layers.assign(input=bbox_pred, output=mask_pred) def process_boxes(): bbox = fluid.layers.slice(bbox_pred, [1], starts=[2], ends=[6]) im_scale = fluid.layers.slice(im_info, [1], starts=[2], ends=[3]) im_scale = fluid.layers.sequence_expand(im_scale, bbox) bbox = fluid.layers.cast(bbox, dtype='float32') im_scale = fluid.layers.cast(im_scale, dtype='float32') mask_rois = bbox * im_scale mask_feat = self.roi_extractor( body_feats, mask_rois, spatial_scale, is_mask=True) if self.with_semantic: semantic_roi_feat = self.semantic_roi_extractor(semantic_feat, mask_rois) if semantic_roi_feat is not None: mask_feat = fluid.layers.sum([mask_feat, semantic_roi_feat]) mask_logits_list = [] mask_pred_list = [] for i in range(self.num_stage): if self.mask_info_flow: last_feat = None for j in range(i): last_feat = self.mask_head.get_output( mask_feat, last_feat, return_logits=False, return_feat=True, wb_scalar=1.0 / self.cascade_rcnn_loss_weight[i] if self.use_bias_scalar else 1.0, name='_' + str(i) + '_' + str(j)) mask_logits = self.mask_head.get_output( mask_feat, last_feat, return_logits=True, return_feat=False, wb_scalar=1.0 / self.cascade_rcnn_loss_weight[i] if self.use_bias_scalar else 1.0, name='_' + str(i)) mask_logits_list.append(mask_logits) else: mask_logits = self.mask_head.get_output( mask_feat, return_logits=True, return_feat=False, name='_' + str(i)) mask_pred_out = self.mask_head.get_prediction(mask_logits, bbox) mask_pred_list.append(mask_pred_out) mask_pred_out = fluid.layers.sum(mask_pred_list) / float( len(mask_pred_list)) fluid.layers.assign(input=mask_pred_out, output=mask_pred) fluid.layers.cond(cond, noop, process_boxes) return mask_pred, bbox_pred def _input_check(self, require_fields, feed_vars): for var in require_fields: assert var in feed_vars, \ "{} has no {} field".format(feed_vars, var) def _decode_box(self, proposals, bbox_pred, curr_stage): rcnn_loc_delta_r = fluid.layers.reshape( bbox_pred, (-1, self.cls_agnostic_bbox_reg, 4)) # only use fg box delta to decode box rcnn_loc_delta_s = fluid.layers.slice( rcnn_loc_delta_r, axes=[1], starts=[1], ends=[2]) refined_bbox = fluid.layers.box_coder( prior_box=proposals, prior_box_var=self.cascade_bbox_reg_weights[curr_stage], target_box=rcnn_loc_delta_s, code_type='decode_center_size', box_normalized=False, axis=1, ) refined_bbox = fluid.layers.reshape(refined_bbox, shape=[-1, 4]) return refined_bbox def _inputs_def(self, image_shape): im_shape = [None] + image_shape # yapf: disable inputs_def = { 'image': {'shape': im_shape, 'dtype': 'float32', 'lod_level': 0}, 'im_info': {'shape': [None, 3], 'dtype': 'float32', 'lod_level': 0}, 'im_id': {'shape': [None, 1], 'dtype': 'int64', 'lod_level': 0}, 'im_shape': {'shape': [None, 3], 'dtype': 'float32', 'lod_level': 0}, 'gt_bbox': {'shape': [None, 4], 'dtype': 'float32', 'lod_level': 1}, 'gt_class': {'shape': [None, 1], 'dtype': 'int32', 'lod_level': 1}, 'is_crowd': {'shape': [None, 1], 'dtype': 'int32', 'lod_level': 1}, 'gt_mask': {'shape': [None, 2], 'dtype': 'float32', 'lod_level': 3}, # polygon coordinates 'semantic': {'shape': [None, 1, None, None], 'dtype': 'int32', 'lod_level': 0}, } # yapf: enable return inputs_def def build_inputs(self, image_shape=[3, None, None], fields=[ 'image', 'im_info', 'im_id', 'gt_bbox', 'gt_class', 'is_crowd', 'gt_mask', 'semantic' ], multi_scale=False, num_scales=-1, use_flip=None, use_dataloader=True, iterable=False, mask_branch=False): inputs_def = self._inputs_def(image_shape) fields = copy.deepcopy(fields) if multi_scale: ms_def, ms_fields = multiscale_def(image_shape, num_scales, use_flip) inputs_def.update(ms_def) fields += ms_fields self.im_info_names = ['image', 'im_info'] + ms_fields if mask_branch: box_fields = ['bbox', 'bbox_flip'] if use_flip else ['bbox'] for key in box_fields: inputs_def[key] = { 'shape': [6], 'dtype': 'float32', 'lod_level': 1 } fields += box_fields feed_vars = OrderedDict([(key, fluid.data( name=key, shape=inputs_def[key]['shape'], dtype=inputs_def[key]['dtype'], lod_level=inputs_def[key]['lod_level'])) for key in fields]) use_dataloader = use_dataloader and not mask_branch loader = fluid.io.DataLoader.from_generator( feed_list=list(feed_vars.values()), capacity=64, use_double_buffer=True, iterable=iterable) if use_dataloader else None return feed_vars, loader def train(self, feed_vars): return self.build(feed_vars, 'train') def eval(self, feed_vars, multi_scale=None, mask_branch=False): if multi_scale: return self.build_multi_scale(feed_vars, mask_branch) return self.build(feed_vars, 'test') def test(self, feed_vars, exclude_nms=False): assert not exclude_nms, "exclude_nms for {} is not support currently".format( self.__class__.__name__) return self.build(feed_vars, 'test')