add vehicle neck and fix bugs

ppcls/arch/__init__.py

-#copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve.
+#copyright (c) 2021 PaddlePaddle Authors. All Rights Reserve.
 #
 #Licensed under the Apache License, Version 2.0 (the "License");
 #you may not use this file except in compliance with the License.
@@ -17,9 +17,7 @@ import importlib
 
 import paddle.nn as nn
 
-from . import backbone
-from . import gears
-
+from . import backbone, gears
 from .backbone import *
 from .gears import build_gear
 from .utils import *
@@ -40,10 +38,10 @@ class RecModel(nn.Layer):
         super().__init__()
         backbone_config = config["Backbone"]
         backbone_name = backbone_config.pop("name")
-        self.backbone = getattr(backbone_name)(**backbone_config)
+        self.backbone = eval(backbone_name)(**backbone_config)
         if "BackboneStopLayer" in config:
-            backbone_stop_layer = config["BackboneStopLayer"]
-            self.backbone.stop_layer(backbone_stop_layer)
+            backbone_stop_layer = config["BackboneStopLayer"]["name"]
+            self.backbone.stop_after(backbone_stop_layer)
 
         if "Neck" in config:
             self.neck = build_gear(config["Neck"])
@@ -55,10 +53,11 @@ class RecModel(nn.Layer):
         else:
             self.head = None
 
-    def forward(self, x):
-        y = self.backbone(x)
+    def forward(self, x, label):
+        x = self.backbone(x)
         if self.neck is not None:
-            y = self.neck(y)
+            x = self.neck(x)
+        y = x
         if self.head is not None:
-            y = self.head(y)
-        return y
+            y = self.head(x, label)
+        return {"features": x, "logits": y}

ppcls/configs/Vehicle/ResNet50.yaml

@@ -21,8 +21,8 @@ Arch:
   Backbone: 
     name: "ResNet50_last_stage_stride1"
     pretrained: True
-  BackboneStoplayer:
-    name: "adaptive_avg_pool2d_1"
+  BackboneStopLayer:
+    name: "adaptive_avg_pool2d_0"
   Neck:
     name: "VehicleNeck"
     in_channels: 2048
@@ -91,7 +91,7 @@ DataLoader:
 
     sampler:
         name: DistributedRandomIdentitySampler
-        batch_size: 64
+        batch_size: 128
         num_instances: 2
         drop_last: False
         shuffle: True
@@ -117,7 +117,7 @@ DataLoader:
               order: ''
     sampler:
         name: DistributedBatchSampler
-        batch_size: 64
+        batch_size: 128
         drop_last: False
         shuffle: False
     loader:

ppcls/configs/Vehicle/ResNet50_ReID.yaml

@@ -25,8 +25,8 @@ Arch:
   Backbone: 
     name: "ResNet50_last_stage_stride1"
     pretrained: True
-  BackboneStoplayer:
-    name: "adaptive_avg_pool2d_1"
+  BackboneStopLayer:
+    name: "adaptive_avg_pool2d_0"
   Neck:
     name: "VehicleNeck"
     in_channels: 2048
@@ -43,9 +43,9 @@ Loss:
   Train:
     - CELoss:
         weight: 1.0
-    - TripletLossV2:
+    - SupConLoss:
         weight: 1.0
-        margin: 0.5
+        views: 2
   Eval:
     - CELoss:
         weight: 1.0
@@ -70,7 +70,7 @@ DataLoader:
     dataset:
         name: "VeriWild"
         image_root: "/work/dataset/VeRI-Wild/images/"
-        cls_label_path: "/work/dataset/VeRI-Wild/train_test_split/debug_train.txt"
+        cls_label_path: "/work/dataset/VeRI-Wild/train_test_split/train_list_start0.txt"
         transform_ops:
           - ResizeImage:
               size: 224
@@ -105,7 +105,7 @@ DataLoader:
     dataset: 
         name: "VeriWild"
         image_root: "/work/dataset/VeRI-Wild/images"
-        cls_label_path: "/work/dataset/VeRI-Wild/train_test_split/debug_test_query.txt"
+        cls_label_path: "/work/dataset/VeRI-Wild/train_test_split/test_3000_id_query.txt"
         transform_ops:
           - ResizeImage:
               size: 224
@@ -128,7 +128,7 @@ DataLoader:
     dataset: 
         name: "VeriWild"
         image_root: "/work/dataset/VeRI-Wild/images"
-        cls_label_path: "/work/dataset/VeRI-Wild/train_test_split/debug_test.txt"
+        cls_label_path: "/work/dataset/VeRI-Wild/train_test_split/test_3000_id.txt"
         transform_ops:
           - ResizeImage:
               size: 224

ppcls/loss/__init__.py

@@ -11,6 +11,7 @@ from .msmloss import MSMLoss
 from .npairsloss import NpairsLoss
 from .trihardloss import TriHardLoss
 from .triplet import TripletLoss, TripletLossV2
+from .supconloss import SupConLoss
 
 
 class CombinedLoss(nn.Layer):

ppcls/loss/supconloss.py

@@ -32,6 +32,7 @@ class SupConLoss(nn.Layer):
         Returns:
             A loss scalar.
         """
+        features = features["features"]
         if self.num_ids is None:
             self.num_ids = int(features.shape[0] / self.views)
 
@@ -104,4 +105,4 @@ class SupConLoss(nn.Layer):
         loss = -(self.temperature / self.base_temperature) * mean_log_prob_pos
         loss = paddle.mean(loss.reshape([anchor_count, batch_size]))
 
-        return loss
+        return {"SupConLoss": loss}