delete useless code

Author: Bin Xiao

README.md

@@ -13,7 +13,7 @@ We start from a high-resolution subnetwork as the first stage, gradually add hig
 | pose_resnet_50     | 96.4 |     95.3 |  89.0 |  83.2 | 88.4 | 84.0 |  79.6 | 88.5 |     34.0 |
 | pose_resnet_101    | 96.9 |     95.9 |  89.5 |  84.4 | 88.4 | 84.5 |  80.7 | 89.1 |     34.0 |
 | pose_resnet_152    | 97.0 |     95.9 |  90.0 |  85.0 | 89.2 | 85.3 |  81.3 | 89.6 |     35.0 |
-| **pose_hrnet_w32** | 97.1 |     95.9 |  90.7 |  86.5 | 89.1 | 87.0 |  83.5 | 90.4 |     37.7 |
+| **pose_hrnet_w32** | 97.1 |     95.9 |  90.3 |  86.4 | 89.1 | 87.1 |  83.3 | 90.3 |     37.7 |
 
 ### Note:
 - Flip test is used.
@@ -181,7 +181,8 @@ ${POSE_ROOT}
 
 ### Training and Testing
 
-#### Testing on MPII dataset using model zoo's models([OneDrive] or [GoogleDrive])
+#### Testing on MPII dataset using model zoo's models([GoogleDrive](https://drive.google.com/drive/folders/1hOTihvbyIxsm5ygDpbUuJ7O_tzv4oXjC?usp=sharing) or [OneDrive](https://1drv.ms/f/s!AhIXJn_J-blW231MH2krnmLq5kkQ))
+ 
 
 ```
 python tools/test.py \
@@ -196,7 +197,8 @@ python tools/train.py \
     --cfg experiments/mpii/hrnet/w32_256x256_adam_lr1e-3.yaml
 ```
 
-#### Testing on COCO val2017 dataset using model zoo's models
+#### Testing on COCO val2017 dataset using model zoo's models([GoogleDrive](https://drive.google.com/drive/folders/1hOTihvbyIxsm5ygDpbUuJ7O_tzv4oXjC?usp=sharing) or [OneDrive](https://1drv.ms/f/s!AhIXJn_J-blW231MH2krnmLq5kkQ))
+ 
 
 ```
 python tools/test.py \

lib/config/models.py

@@ -21,33 +21,12 @@
 POSE_RESNET.FINAL_CONV_KERNEL = 1
 POSE_RESNET.PRETRAINED_LAYERS = ['*']
 
-# pose_resnet related params
-POSE_RESNEXT = CN()
-POSE_RESNEXT.NUM_LAYERS = 50
-POSE_RESNEXT.DECONV_WITH_BIAS = False
-POSE_RESNEXT.NUM_DECONV_LAYERS = 3
-POSE_RESNEXT.NUM_DECONV_FILTERS = [256, 256, 256]
-POSE_RESNEXT.NUM_DECONV_KERNELS = [4, 4, 4]
-POSE_RESNEXT.FINAL_CONV_KERNEL = 1
-POSE_RESNEXT.BASE_WIDTH = 4
-POSE_RESNEXT.CARDINALITY = 32
-POSE_RESNEXT.STEM_KERNEL_SIZE = 7
-POSE_RESNEXT.PRETRAINED_LAYERS = ['*']
-
 # pose_multi_resoluton_net related params
 POSE_HIGH_RESOLUTION_NET = CN()
 POSE_HIGH_RESOLUTION_NET.PRETRAINED_LAYERS = ['*']
 POSE_HIGH_RESOLUTION_NET.STEM_INPLANES = 64
 POSE_HIGH_RESOLUTION_NET.FINAL_CONV_KERNEL = 1
 
-# POSE_HIGH_RESOLUTION_NET.STAGE1 = CN()
-# POSE_HIGH_RESOLUTION_NET.STAGE1.NUM_MODULES = 1
-# POSE_HIGH_RESOLUTION_NET.STAGE1.NUM_BRANCHES = 1
-# POSE_HIGH_RESOLUTION_NET.STAGE1.NUM_BLOCKS = [4]
-# POSE_HIGH_RESOLUTION_NET.STAGE1.NUM_CHANNELS = [64]
-# POSE_HIGH_RESOLUTION_NET.STAGE1.BLOCK = 'BOTTLENECK'
-# POSE_HIGH_RESOLUTION_NET.STAGE1.FUSE_METHOD = 'CAT'
-
 POSE_HIGH_RESOLUTION_NET.STAGE2 = CN()
 POSE_HIGH_RESOLUTION_NET.STAGE2.NUM_MODULES = 1
 POSE_HIGH_RESOLUTION_NET.STAGE2.NUM_BRANCHES = 2
@@ -75,7 +54,5 @@
 
 MODEL_EXTRAS = {
     'pose_resnet': POSE_RESNET,
-    'pose_resnetstem3x3': POSE_RESNET,
-    'pose_resNeXt': POSE_RESNEXT,
     'pose_high_resolution_net': POSE_HIGH_RESOLUTION_NET,
 }

lib/dataset/__init__.py

@@ -8,9 +8,5 @@
 from __future__ import division
 from __future__ import print_function
 
-from __future__ import absolute_import
-from __future__ import division
-from __future__ import print_function
-
 from .mpii import MPIIDataset as mpii
 from .coco import COCODataset as coco