[FasterRCNN] add sample patches to tensorboard

examples/FasterRCNN/README.md

@@ -47,7 +47,7 @@ To evaluate the performance (pretrained models can be downloaded in [model zoo](
 Mean Average Precision @IoU=0.50:0.95:
 
 + trainval35k/minival, FASTRCNN_BATCH=256: 34.2. Takes 49h on 8 TitanX.
-+ trainval35k/minival, FASTRCNN_BATCH=64: 32.7. Takes 31h on 8 TitanX.
++ trainval35k/minival, FASTRCNN_BATCH=64: 32.7. Takes 25h on 8 TitanX.
 
 The hyperparameters are not carefully tuned. You can probably get better performance by e.g. training longer.
 

examples/FasterRCNN/train.py

@@ -25,7 +25,7 @@ from coco import COCODetection
 from basemodel import (
     image_preprocess, pretrained_resnet_conv4, resnet_conv5)
 from model import (
-    clip_boxes, decode_bbox_target, encode_bbox_target,
+    clip_boxes, decode_bbox_target, encode_bbox_target, crop_and_resize,
     rpn_head, rpn_losses,
     generate_rpn_proposals, sample_fast_rcnn_targets, roi_align,
     fastrcnn_head, fastrcnn_losses, fastrcnn_predictions)
@@ -81,7 +81,7 @@ class Model(ModelDesc):
         is_training = get_current_tower_context().is_training
         image, anchor_labels, anchor_boxes, gt_boxes, gt_labels = inputs
         fm_anchors = self._get_anchors(image)
-        image = self._preprocess(image)
+        image = self._preprocess(image)     # 1CHW
         image_shape2d = tf.shape(image)[2:]
 
         anchor_boxes_encoded = encode_bbox_target(anchor_boxes, fm_anchors)
@@ -116,6 +116,13 @@ class Model(ModelDesc):
             fg_inds_wrt_sample = tf.reshape(tf.where(rcnn_labels > 0), [-1])   # fg inds w.r.t all samples
             fg_sampled_boxes = tf.gather(rcnn_sampled_boxes, fg_inds_wrt_sample)
 
+            with tf.name_scope('fg_sample_patch_viz'):
+                fg_sampled_patches = crop_and_resize(
+                    image, fg_sampled_boxes,
+                    tf.zeros_like(fg_inds_wrt_sample, dtype=tf.int32), [300, 300])
+                fg_sampled_patches = tf.transpose(fg_sampled_patches, [0, 2, 3, 1])
+                tf.summary.image('viz', fg_sampled_patches, max_outputs=30)
+
             matched_gt_boxes = tf.gather(gt_boxes, fg_inds_wrt_gt)
             encoded_boxes = encode_bbox_target(
                 matched_gt_boxes,