Fbnet benchmark

MODEL_ZOO.md

@@ -33,7 +33,28 @@ backbone | type | lr sched | im / gpu | train mem(GB) | train time (s/iter) | to
 -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | --
 R-50-FPN | Keypoint | 1x | 2 | 5.7 | 0.3771 | 9.4 | 0.10941 | 53.7 | 64.3 | 9981060
 
+### Light-weight Model baselines
 
+We provided pre-trained models for selected FBNet models. 
+* All the models are trained from scratched with BN using the training schedule specified below. 
+* Evaluation is performed on a single NVIDIA V100 GPU with `MODEL.RPN.POST_NMS_TOP_N_TEST` set to `200`. 
+
+The following inference time is reported:
+  * inference total batch=8: Total inference time including data loading, model inference and pre/post preprocessing using 8 images per batch.
+  * inference model batch=8: Model inference time only and using 8 images per batch.
+  * inference model batch=1: Model inference time only and using 1 image per batch.
+  * inferenee caffe2 batch=1: Model inference time for the model in Caffe2 format using 1 image per batch. The Caffe2 models fused the BN to Conv and purely run on C++/CUDA by using Caffe2 ops for rpn/detection post processing.
+
+The pre-trained models are available in the link in the model id.
+
+backbone | type | resolution | lr sched | im / gpu | train mem(GB) | train time (s/iter) | total train time (hr) | inference total batch=8 (s/im) | inference model batch=8 (s/im) | inference model batch=1 (s/im) | inference caffe2 batch=1 (s/im) | box AP | mask AP | model id
+-- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | -- | --
+[R-50-C4](configs/e2e_faster_rcnn_R_50_C4_1x.yaml) (reference) | Fast | 800 | 1x | 1 | 5.8 | 0.4036 | 20.2 | 0.0875 | **0.0793** | 0.0831 | **0.0625** | 34.4 | - | f35857197
+[fbnet_chamv1a](configs/e2e_faster_rcnn_fbnet_chamv1a_600.yaml) | Fast | 600 | 0.75x | 12 | 13.6 | 0.5444 | 20.5 | 0.0315 | **0.0260** | 0.0376 | **0.0188** | 33.5 | - | f100940543
+[fbnet_default](configs/e2e_faster_rcnn_fbnet_600.yaml) | Fast | 600 | 0.5x | 16 | 11.1 | 0.4872 | 12.5 | 0.0316 | **0.0250** | 0.0297 | **0.0130** | 28.2 | - | f101086388
+[R-50-C4](configs/e2e_mask_rcnn_R_50_C4_1x.yaml) (reference) | Mask | 800 | 1x | 1 | 5.8 | 0.452 | 22.6 | 0.0918 | **0.0848** | 0.0844 | - | 35.2 | 31.0 | f35858791
+[fbnet_xirb16d](configs/e2e_mask_rcnn_fbnet_xirb16d_dsmask_600.yaml) | Mask | 600 | 0.5x | 16 | 13.4 | 1.1732 | 29 | 0.0386 | **0.0319** | 0.0356 | - | 30.7 | 26.9 | f101086394
+[fbnet_default](configs/e2e_mask_rcnn_fbnet_600.yaml) | Mask | 600 | 0.5x | 16 | 13.0 | 0.9036 | 23.0 | 0.0327 | **0.0269** | 0.0385 | - | 29.0 | 26.1 | f101086385
 
 ## Comparison with Detectron and mmdetection
 

configs/e2e_faster_rcnn_fbnet.yaml

@@ -15,7 +15,7 @@ MODEL:
     PRE_NMS_TOP_N_TRAIN: 6000
     PRE_NMS_TOP_N_TEST: 6000
     POST_NMS_TOP_N_TRAIN: 2000
-    POST_NMS_TOP_N_TEST: 1000
+    POST_NMS_TOP_N_TEST: 100
     RPN_HEAD: FBNet.rpn_head
   ROI_HEADS:
     BATCH_SIZE_PER_IMAGE: 512

configs/e2e_faster_rcnn_fbnet_600.yaml

@@ -15,7 +15,7 @@ MODEL:
     PRE_NMS_TOP_N_TRAIN: 6000
     PRE_NMS_TOP_N_TEST: 6000
     POST_NMS_TOP_N_TRAIN: 2000
-    POST_NMS_TOP_N_TEST: 1000
+    POST_NMS_TOP_N_TEST: 200
     RPN_HEAD: FBNet.rpn_head
   ROI_HEADS:
     BATCH_SIZE_PER_IMAGE: 256

configs/e2e_faster_rcnn_fbnet_chamv1a_600.yaml

@@ -0,0 +1,44 @@
+MODEL:
+  META_ARCHITECTURE: "GeneralizedRCNN"
+  BACKBONE:
+    CONV_BODY: FBNet
+  FBNET:
+    ARCH: "cham_v1a"
+    BN_TYPE: "bn"
+    WIDTH_DIVISOR: 8
+    DW_CONV_SKIP_BN: True
+    DW_CONV_SKIP_RELU: True
+  RPN:
+    ANCHOR_SIZES: (32, 64, 128, 256, 512)
+    ANCHOR_STRIDE: (16, )
+    BATCH_SIZE_PER_IMAGE: 256
+    PRE_NMS_TOP_N_TRAIN: 6000
+    PRE_NMS_TOP_N_TEST: 6000
+    POST_NMS_TOP_N_TRAIN: 2000
+    POST_NMS_TOP_N_TEST: 200
+    RPN_HEAD: FBNet.rpn_head
+  ROI_HEADS:
+    BATCH_SIZE_PER_IMAGE: 128
+  ROI_BOX_HEAD:
+    POOLER_RESOLUTION: 6
+    FEATURE_EXTRACTOR: FBNet.roi_head
+    NUM_CLASSES: 81
+DATASETS:
+  TRAIN: ("coco_2014_train", "coco_2014_valminusminival")
+  TEST: ("coco_2014_minival",)
+SOLVER:
+  BASE_LR: 0.045
+  WARMUP_FACTOR: 0.1
+  WEIGHT_DECAY: 0.0001
+  STEPS: (90000, 120000)
+  MAX_ITER: 135000
+  IMS_PER_BATCH: 96  # for 8GPUs
+# TEST:
+#   IMS_PER_BATCH: 8
+INPUT:
+  MIN_SIZE_TRAIN: (600, )
+  MAX_SIZE_TRAIN: 1000
+  MIN_SIZE_TEST: 600
+  MAX_SIZE_TEST: 1000
+  PIXEL_MEAN: [103.53, 116.28, 123.675]
+  PIXEL_STD: [57.375, 57.12, 58.395]

configs/e2e_mask_rcnn_fbnet.yaml

@@ -8,15 +8,15 @@ MODEL:
     WIDTH_DIVISOR: 8
     DW_CONV_SKIP_BN: True
     DW_CONV_SKIP_RELU: True
-    DET_HEAD_LAST_SCALE: -1.0
+    DET_HEAD_LAST_SCALE: 0.0
   RPN:
     ANCHOR_SIZES: (16, 32, 64, 128, 256)
     ANCHOR_STRIDE: (16, )
     BATCH_SIZE_PER_IMAGE: 256
     PRE_NMS_TOP_N_TRAIN: 6000
     PRE_NMS_TOP_N_TEST: 6000
     POST_NMS_TOP_N_TRAIN: 2000
-    POST_NMS_TOP_N_TEST: 1000
+    POST_NMS_TOP_N_TEST: 100
     RPN_HEAD: FBNet.rpn_head
   ROI_HEADS:
     BATCH_SIZE_PER_IMAGE: 256

configs/e2e_mask_rcnn_fbnet_600.yaml

@@ -0,0 +1,52 @@
+MODEL:
+  META_ARCHITECTURE: "GeneralizedRCNN"
+  BACKBONE:
+    CONV_BODY: FBNet
+  FBNET:
+    ARCH: "default"
+    BN_TYPE: "bn"
+    WIDTH_DIVISOR: 8
+    DW_CONV_SKIP_BN: True
+    DW_CONV_SKIP_RELU: True
+    DET_HEAD_LAST_SCALE: 0.0
+  RPN:
+    ANCHOR_SIZES: (32, 64, 128, 256, 512)
+    ANCHOR_STRIDE: (16, )
+    BATCH_SIZE_PER_IMAGE: 256
+    PRE_NMS_TOP_N_TRAIN: 6000
+    PRE_NMS_TOP_N_TEST: 6000
+    POST_NMS_TOP_N_TRAIN: 2000
+    POST_NMS_TOP_N_TEST: 200
+    RPN_HEAD: FBNet.rpn_head
+  ROI_HEADS:
+    BATCH_SIZE_PER_IMAGE: 256
+  ROI_BOX_HEAD:
+    POOLER_RESOLUTION: 6
+    FEATURE_EXTRACTOR: FBNet.roi_head
+    NUM_CLASSES: 81
+  ROI_MASK_HEAD:
+    POOLER_RESOLUTION: 6
+    FEATURE_EXTRACTOR: FBNet.roi_head_mask
+    PREDICTOR: "MaskRCNNConv1x1Predictor"
+    RESOLUTION: 12
+    SHARE_BOX_FEATURE_EXTRACTOR: False
+  MASK_ON: True
+DATASETS:
+  TRAIN: ("coco_2014_train", "coco_2014_valminusminival")
+  TEST: ("coco_2014_minival",)
+SOLVER:
+  BASE_LR: 0.06
+  WARMUP_FACTOR: 0.1
+  WEIGHT_DECAY: 0.0001
+  STEPS: (60000, 80000)
+  MAX_ITER: 90000
+  IMS_PER_BATCH: 128  # for 8GPUs
+# TEST:
+#   IMS_PER_BATCH: 8
+INPUT:
+  MIN_SIZE_TRAIN: (600, )
+  MAX_SIZE_TRAIN: 1000
+  MIN_SIZE_TEST: 600
+  MAX_SIZE_TEST: 1000
+  PIXEL_MEAN: [103.53, 116.28, 123.675]
+  PIXEL_STD: [57.375, 57.12, 58.395]

configs/e2e_mask_rcnn_fbnet_xirb16d_dsmask.yaml

@@ -16,7 +16,7 @@ MODEL:
     PRE_NMS_TOP_N_TRAIN: 6000
     PRE_NMS_TOP_N_TEST: 6000
     POST_NMS_TOP_N_TRAIN: 2000
-    POST_NMS_TOP_N_TEST: 1000
+    POST_NMS_TOP_N_TEST: 100
     RPN_HEAD: FBNet.rpn_head
   ROI_HEADS:
     BATCH_SIZE_PER_IMAGE: 512

configs/e2e_mask_rcnn_fbnet_xirb16d_dsmask_600.yaml

@@ -0,0 +1,52 @@
+MODEL:
+  META_ARCHITECTURE: "GeneralizedRCNN"
+  BACKBONE:
+    CONV_BODY: FBNet
+  FBNET:
+    ARCH: "xirb16d_dsmask"
+    BN_TYPE: "bn"
+    WIDTH_DIVISOR: 8
+    DW_CONV_SKIP_BN: True
+    DW_CONV_SKIP_RELU: True
+    DET_HEAD_LAST_SCALE: 0.0
+  RPN:
+    ANCHOR_SIZES: (32, 64, 128, 256, 512)
+    ANCHOR_STRIDE: (16, )
+    BATCH_SIZE_PER_IMAGE: 256
+    PRE_NMS_TOP_N_TRAIN: 6000
+    PRE_NMS_TOP_N_TEST: 6000
+    POST_NMS_TOP_N_TRAIN: 2000
+    POST_NMS_TOP_N_TEST: 200
+    RPN_HEAD: FBNet.rpn_head
+  ROI_HEADS:
+    BATCH_SIZE_PER_IMAGE: 256
+  ROI_BOX_HEAD:
+    POOLER_RESOLUTION: 6
+    FEATURE_EXTRACTOR: FBNet.roi_head
+    NUM_CLASSES: 81
+  ROI_MASK_HEAD:
+    POOLER_RESOLUTION: 6
+    FEATURE_EXTRACTOR: FBNet.roi_head_mask
+    PREDICTOR: "MaskRCNNConv1x1Predictor"
+    RESOLUTION: 12
+    SHARE_BOX_FEATURE_EXTRACTOR: False
+  MASK_ON: True
+DATASETS:
+  TRAIN: ("coco_2014_train", "coco_2014_valminusminival")
+  TEST: ("coco_2014_minival",)
+SOLVER:
+  BASE_LR: 0.06
+  WARMUP_FACTOR: 0.1
+  WEIGHT_DECAY: 0.0001
+  STEPS: (60000, 80000)
+  MAX_ITER: 90000
+  IMS_PER_BATCH: 128  # for 8GPUs
+# TEST:
+#   IMS_PER_BATCH: 8
+INPUT:
+  MIN_SIZE_TRAIN: (600, )
+  MAX_SIZE_TRAIN: 1000
+  MIN_SIZE_TEST: 600
+  MAX_SIZE_TEST: 1000
+  PIXEL_MEAN: [103.53, 116.28, 123.675]
+  PIXEL_STD: [57.375, 57.12, 58.395]

maskrcnn_benchmark/engine/inference.py

@@ -1,5 +1,4 @@
 # Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
-import datetime
 import logging
 import time
 import os
@@ -11,17 +10,23 @@
 from ..utils.comm import is_main_process
 from ..utils.comm import all_gather
 from ..utils.comm import synchronize
+from ..utils.timer import Timer, get_time_str
 
 
-def compute_on_dataset(model, data_loader, device):
+def compute_on_dataset(model, data_loader, device, timer=None):
     model.eval()
     results_dict = {}
     cpu_device = torch.device("cpu")
-    for i, batch in enumerate(tqdm(data_loader)):
+    for _, batch in enumerate(tqdm(data_loader)):
         images, targets, image_ids = batch
         images = images.to(device)
         with torch.no_grad():
+            if timer:
+                timer.tic()
             output = model(images)
+            if timer:
+                torch.cuda.synchronize()
+                timer.toc()
             output = [o.to(cpu_device) for o in output]
         results_dict.update(
             {img_id: result for img_id, result in zip(image_ids, output)}
@@ -72,17 +77,27 @@ def inference(
     logger = logging.getLogger("maskrcnn_benchmark.inference")
     dataset = data_loader.dataset
     logger.info("Start evaluation on {} dataset({} images).".format(dataset_name, len(dataset)))
-    start_time = time.time()
-    predictions = compute_on_dataset(model, data_loader, device)
+    total_timer = Timer()
+    inference_timer = Timer()
+    total_timer.tic()
+    predictions = compute_on_dataset(model, data_loader, device, inference_timer)
     # wait for all processes to complete before measuring the time
     synchronize()
-    total_time = time.time() - start_time
-    total_time_str = str(datetime.timedelta(seconds=total_time))
+    total_time = total_timer.toc()
+    total_time_str = get_time_str(total_time)
     logger.info(
-        "Total inference time: {} ({} s / img per device, on {} devices)".format(
+        "Total run time: {} ({} s / img per device, on {} devices)".format(
             total_time_str, total_time * num_devices / len(dataset), num_devices
         )
     )
+    total_infer_time = get_time_str(inference_timer.total_time)
+    logger.info(
+        "Model inference time: {} ({} s / img per device, on {} devices)".format(
+            total_infer_time,
+            inference_timer.total_time * num_devices / len(dataset),
+            num_devices,
+        )
+    )
 
     predictions = _accumulate_predictions_from_multiple_gpus(predictions)
     if not is_main_process():

maskrcnn_benchmark/modeling/backbone/fbnet.py

@@ -199,13 +199,6 @@ def __init__(
             ("last", last)
         ]))
 
-        # output_blob = builder.add_final_pool(
-        #     # model, output_blob, kernel_size=cfg.FAST_RCNN.ROI_XFORM_RESOLUTION)
-        #     model,
-        #     output_blob,
-        #     kernel_size=int(cfg.FAST_RCNN.ROI_XFORM_RESOLUTION / stride_init),
-        # )
-
         self.out_channels = builder.last_depth
 
     def forward(self, x, proposals):

maskrcnn_benchmark/modeling/backbone/fbnet_builder.py

@@ -771,6 +771,9 @@ def add_last(self, stage_info):
             last_channel = int(self.last_depth * (-channel_scale))
         last_channel = self._get_divisible_width(last_channel)
 
+        if last_channel == 0:
+            return nn.Sequential()
+
         dim_in = self.last_depth
         ret = ConvBNRelu(
             dim_in,