add new model

0d97cc8c · Sugon_ldc · 0d97cc8c · 0d97cc8c · 0d97cc8c · 0d97cc8c
Commit 0d97cc8c authored Jun 07, 2023 by Sugon_ldc
20 changed files
--- a/configs/deeplabv3p/deeplabv3p_resnet50_os8_voc12aug_512x512_40k.yml
+++ b/configs/deeplabv3p/deeplabv3p_resnet50_os8_voc12aug_512x512_40k.yml
+_base_: '../_base_/pascal_voc12aug.yml'
+
+model:
+  type: DeepLabV3P
+  backbone:
+    type: ResNet50_vd
+    output_stride: 8
+    multi_grid: [1, 2, 4]
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet50_vd_ssld_v2.tar.gz
+  backbone_indices: [0, 3]
+  aspp_ratios: [1, 12, 24, 36]
+  aspp_out_channels: 256
+  align_corners: False
+  pretrained: null
--- a/configs/dmnet/README.md
+++ b/configs/dmnet/README.md
+# Dynamic Multi-scale Filters for Semantic Segmentation
+
+## Reference
+
+> Junjun He, Zhongying Deng, Yu Qiao. "Dynamic Multi-scale Filters for Semantic Segmentation" In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 3562-3572. 2019.
+
+## Performance
+
+### Cityscapes
+
+| Model | Backbone | Resolution | Training Iters | mIoU | mIoU (flip) | mIoU (ms+flip) | Links |
+|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|
+|DMNet|ResNet101_vd|1024x512|80000|79.76%|80.11%|80.56%|[model](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/dmnet_resnet101_os8_cityscapes_1024x512_80k/model.pdparams) \| [log](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/dmnet_resnet101_os8_cityscapes_1024x512_80k/train.log) \| [vdl](https://www.paddlepaddle.org.cn/paddle/visualdl/service/app/scalar?id=d5bac108e3ff90136771b677d8459d17)
--- a/configs/dmnet/dmnet_resnet101_os8_cityscapes_1024x512_80k.yml
+++ b/configs/dmnet/dmnet_resnet101_os8_cityscapes_1024x512_80k.yml
+_base_: '../_base_/cityscapes.yml'
+
+batch_size: 2
+iters: 80000
+
+model:
+  type: DMNet
+  backbone:
+    type: ResNet101_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet101_vd_ssld.tar.gz
+
+optimizer:
+  type: sgd
+  weight_decay: 0.0005
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
+
+lr_scheduler:
+  type: PolynomialDecay
+  learning_rate: 0.01
+  end_lr: 0.0
+  power: 0.9
--- a/configs/dnlnet/README.md
+++ b/configs/dnlnet/README.md
+# Disentangled Non-Local Neural Networks
+
+## Reference
+
+> Minghao Yin, Zhuliang Yao, Yue Cao, Xiu Li, Zheng Zhang, Stephen Lin, Han Hu:
+Disentangled Non-local Neural Networks. ECCV (15) 2020: 191-207.
+
+## Performance
+
+### Cityscapes
+
+| Model | Backbone | Resolution | Training Iters | mIoU | mIoU (flip) | mIoU (ms+flip) |Links |
+|-|-|-|-|-|-|-|-|
+|DNLNet|ResNet50_OS8|1024x512|80000|79.95%|80.43%|-|[model](https://paddleseg.bj.bcebos.com/dygraph/cityscapes/dnlnet_resnet50_os8_cityscapes_1024x512_80k/model.pdparams) \| [log](https://paddleseg.bj.bcebos.com/dygraph/cityscapes/dnlnet_resnet50_os8_cityscapes_1024x512_80k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=922cf0682c5e684507ab54a14ef12847)|
+|DNLNet|ResNet101_OS8|1024x512|80000|81.03%|81.38%|-|[model](https://paddleseg.bj.bcebos.com/dygraph/cityscapes/dnlnet_resnet101_os8_cityscapes_1024x512_80k/model.pdparams) \| [log](https://paddleseg.bj.bcebos.com/dygraph/cityscapes/dnlnet_resnet101_os8_cityscapes_1024x512_80k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=3e0d13c4d9dbf4115bbba2abdc88122c)|
+
+### Pascal VOC 2012 + Aug
+
+| Model | Backbone | Resolution | Training Iters | mIoU | mIoU (flip) | mIoU (ms+flip) | Links |
+|-|-|-|-|-|-|-|-|
+|DNLNet|ResNet50_OS8|512x512|40000|80.89%|81.31%|81.56%|[model](https://paddleseg.bj.bcebos.com/dygraph/pascal_voc12/dnlnet_resnet50_os8_voc12aug_512x512_40k/model.pdparams) \| [log](https://paddleseg.bj.bcebos.com/dygraph/pascal_voc12/dnlnet_resnet50_os8_voc12aug_512x512_40k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=8877c77bef8b227af22c5eb3017138ce)|
+|DNLNet|ResNet101_OS8|512x512|40000|80.49%|80.83%| 81.33%|[model](https://paddleseg.bj.bcebos.com/dygraph/pascal_voc12/dnlnet_resnet101_os8_voc12aug_512x512_40k/model.pdparams) \| [log](https://paddleseg.bj.bcebos.com/dygraph/pascal_voc12/dnlnet_resnet101_os8_voc12aug_512x512_40k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=1d42c22da1c465d9a38e4204bebeeb54)|
--- a/configs/dnlnet/dnlnet_resnet101_os8_cityscapes_1024x512_80k.yml
+++ b/configs/dnlnet/dnlnet_resnet101_os8_cityscapes_1024x512_80k.yml
+_base_: '../_base_/cityscapes.yml'
+
+batch_size: 2
+iters: 80000
+
+model:
+  type: DNLNet
+  backbone:
+    type: ResNet101_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet101_vd_ssld.tar.gz
+  num_classes: 19
+
+optimizer:
+  type: sgd
+  momentum: 0.9
+  weight_decay: 0.00004
+
+lr_scheduler:
+  type: PolynomialDecay
+  learning_rate: 0.01
+  power: 0.9
+
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
--- a/configs/dnlnet/dnlnet_resnet101_os8_voc12aug_512x512_40k.yml
+++ b/configs/dnlnet/dnlnet_resnet101_os8_voc12aug_512x512_40k.yml
+_base_: '../_base_/pascal_voc12aug.yml'
+
+model:
+  type: DNLNet
+  backbone:
+    type: ResNet101_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet101_vd_ssld.tar.gz
+
+optimizer:
+  type: sgd
+  momentum: 0.9
+  weight_decay: 4.0e-05
+
+lr_scheduler:
+  type: PolynomialDecay
+  learning_rate: 0.01
+  power: 0.9
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
--- a/configs/dnlnet/dnlnet_resnet50_os8_cityscapes_1024x512_80k.yml
+++ b/configs/dnlnet/dnlnet_resnet50_os8_cityscapes_1024x512_80k.yml
+_base_: '../_base_/cityscapes.yml'
+
+batch_size: 2
+iters: 80000
+
+model:
+  type: DNLNet
+  backbone:
+    type: ResNet50_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet50_vd_ssld_v2.tar.gz
+  num_classes: 19
+
+optimizer:
+  type: sgd
+  momentum: 0.9
+  weight_decay: 0.00004
+
+lr_scheduler:
+  type: PolynomialDecay
+  learning_rate: 0.01
+  power: 0.9
+
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
--- a/configs/dnlnet/dnlnet_resnet50_os8_voc12aug_512x512_40k.yml
+++ b/configs/dnlnet/dnlnet_resnet50_os8_voc12aug_512x512_40k.yml
+_base_: '../_base_/pascal_voc12aug.yml'
+
+model:
+  type: DNLNet
+  backbone:
+    type: ResNet50_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet50_vd_ssld_v2.tar.gz
+
+optimizer:
+  type: sgd
+  momentum: 0.9
+  weight_decay: 4.0e-05
+
+lr_scheduler:
+  type: PolynomialDecay
+  learning_rate: 0.01
+  power: 0.9
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
--- a/configs/emanet/README.md
+++ b/configs/emanet/README.md
+# Expectation-Maximization Attention Networks for Semantic Segmentation
+
+## Reference
+
+> Xia Li, Zhisheng Zhong, Jianlong Wu, Yibo Yang, Zhouchen Lin, Hong Liu:
+Expectation-Maximization Attention Networks for Semantic Segmentation. ICCV 2019: 9166-9175.
+
+## Performance
+
+### Cityscapes
+
+| Model | Backbone | Resolution | Training Iters | mIoU | mIoU (flip) | mIoU (ms+flip) |Links |
+|-|-|-|-|-|-|-|-|
+|EMANet|ResNet50_OS8|1024x512|80000|79.05%|79.34%|79.69%|[model](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/emanet_resnet50_os8_cityscapes_1024x512_80k/model.pdparams) \| [log](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/emanet_resnet50_os8_cityscapes_1024x512_80k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=0a05a0c4cd7d785b9707bdc59f55f585)|
+|EMANet|ResNet101_OS8|1024x512|80000|80.00%|80.23%|80.53%|[model](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/emanet_resnet101_os8_cityscapes_1024x512_80k/model.pdparams) \| [log](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/emanet_resnet101_os8_cityscapes_1024x512_80k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=ee6926322b8e292ce23ce62ecdaa3439)|
+
+### Pascal VOC 2012 + Aug
+
+| Model | Backbone | Resolution | Training Iters | mIoU | mIoU (flip) | mIoU (ms+flip) | Links |
+|-|-|-|-|-|-|-|-|
+|EMANet|ResNet50_OS8|512x512|40000|78.60%|78.90%|79.17%|[model](https://bj.bcebos.com/paddleseg/dygraph/pascal_voc12/emanet_resnet50_os8_voc12aug_512x512_40k/model.pdparams) \| [log](https://bj.bcebos.com/paddleseg/dygraph/pascal_voc12/emanet_resnet50_os8_voc12aug_512x512_40k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=3e60b80b984a71f3d2b83b8a746a819c)|
+|EMANet|ResNet101_OS8|512x512|40000|79.47%|79.97%| 80.67%|[model](https://bj.bcebos.com/paddleseg/dygraph/pascal_voc12/emanet_resnet101_os8_voc12aug_512x512_40k/model.pdparams) \| [log](https://bj.bcebos.com/paddleseg/dygraph/pascal_voc12/emanet_resnet101_os8_voc12aug_512x512_40k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=f33479772409766dbc40b5f031cbdb1a)|
--- a/configs/emanet/emanet_resnet101_os8_cityscapes_1024x512_80k.yml
+++ b/configs/emanet/emanet_resnet101_os8_cityscapes_1024x512_80k.yml
+_base_: '../_base_/cityscapes.yml'
+
+batch_size: 2
+iters: 80000
+
+model:
+  type: EMANet
+  backbone:
+    type: ResNet101_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet101_vd_ssld.tar.gz
+  num_classes: 19
+  ema_channels: 512
+  gc_channels: 256
+  num_bases: 64
+  stage_num: 3
+  momentum: 0.1
+  concat_input: True
+  enable_auxiliary_loss: True
+  align_corners: False
+
+optimizer:
+  type: sgd
+  momentum: 0.9
+  weight_decay: 0.0005
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
--- a/configs/emanet/emanet_resnet101_os8_voc12aug_512x512_40k.yml
+++ b/configs/emanet/emanet_resnet101_os8_voc12aug_512x512_40k.yml
+_base_: '../_base_/pascal_voc12aug.yml'
+
+model:
+  type: EMANet
+  backbone:
+    type: ResNet101_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet101_vd_ssld.tar.gz
+  ema_channels: 512
+  gc_channels: 256
+  num_bases: 64
+  stage_num: 3
+  momentum: 0.1
+  concat_input: True
+  enable_auxiliary_loss: True
+  align_corners: True
+
+optimizer:
+  type: sgd
+  momentum: 0.9
+  weight_decay: 0.0005
+
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
--- a/configs/emanet/emanet_resnet50_os8_cityscapes_1024x512_80k.yml
+++ b/configs/emanet/emanet_resnet50_os8_cityscapes_1024x512_80k.yml
+_base_: '../_base_/cityscapes.yml'
+
+batch_size: 2
+iters: 80000
+
+model:
+  type: EMANet
+  backbone:
+    type: ResNet50_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet50_vd_ssld_v2.tar.gz
+  num_classes: 19
+  ema_channels: 512
+  gc_channels: 256
+  num_bases: 64
+  stage_num: 3
+  momentum: 0.1
+  concat_input: True
+  enable_auxiliary_loss: True
+  align_corners: False
+
+optimizer:
+  type: sgd
+  momentum: 0.9
+  weight_decay: 0.0005
+
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
--- a/configs/emanet/emanet_resnet50_os8_voc12aug_512x512_40k.yml
+++ b/configs/emanet/emanet_resnet50_os8_voc12aug_512x512_40k.yml
+_base_: '../_base_/pascal_voc12aug.yml'
+
+
+model:
+  type: EMANet
+  backbone:
+    type: ResNet50_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet50_vd_ssld_v2.tar.gz
+  ema_channels: 512
+  gc_channels: 256
+  num_bases: 64
+  stage_num: 3
+  momentum: 0.1
+  concat_input: True
+  enable_auxiliary_loss: True
+  align_corners: True
+
+optimizer:
+  type: sgd
+  momentum: 0.9
+  weight_decay: 0.0005
+
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+  coef: [1, 0.4]
--- a/configs/encnet/README.md
+++ b/configs/encnet/README.md
+# ENCNet: Context Encoding for Semantic Segmentation
+
+## Reference
+> Hang Zhang, Kristin Dana, et, al. "Context Encoding for Semantic Segmentation". In Proceedings of the IEEE conference on Computer Vision and Pattern Recognition, pp. 7151-7160. 2018.
+
+## Performance
+
+### Cityscapes
+
+| Model | Backbone | Resolution | Training Iters | mIoU | mIoU (flip) | mIoU (ms+flip) | Links |
+|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|
+|ENCNet|ResNet101_vd|1024x512|80000|79.42%|80.02%|-|[model](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/encnet_resnet101_os8_cityscapes_1024x512_80k/model.pdparams) \| [log](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/encnet_resnet101_os8_cityscapes_1024x512_80k/train.log )\| [vdl](https://www.paddlepaddle.org.cn/paddle/visualdl/service/app/index?id=c2b819e6b666e4e50bba4b525f515d41)|
--- a/configs/encnet/encnet_resnet101_os8_cityscapes_1024x512_80k.yml
+++ b/configs/encnet/encnet_resnet101_os8_cityscapes_1024x512_80k.yml
+_base_: '../_base_/cityscapes.yml'
+
+batch_size: 2
+iters: 80000
+
+model:
+  type: ENCNet
+  backbone:
+    type: ResNet101_vd
+    output_stride: 8
+    pretrained: https://bj.bcebos.com/paddleseg/dygraph/resnet101_vd_ssld.tar.gz
+  num_codes: 32
+  mid_channels: 512
+  backbone_indices: [1, 2, 3]
+  use_se_loss: True
+  add_lateral: True
+
+optimizer:
+  type: sgd
+  weight_decay: 0.0005
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+    - type: CrossEntropyLoss
+    - type: SECrossEntropyLoss
+  coef: [1, 0.4, 0.2]
+
+lr_scheduler:
+  type: PolynomialDecay
+  learning_rate: 0.01
+  end_lr: 0.0
+  power: 0.9
--- a/configs/enet/README.md
+++ b/configs/enet/README.md
+# ENet: A Deep Neural Network Architecture forReal-Time Semantic Segmentation
+
+## Reference
+> Abhishek Chaurasia, Sangpil Kim, Eugenio Culurciello. "ENet: A Deep Neural Network Architecture for
+Real-Time Semantic Segmentation." arXiv preprint arXiv:1606.02147(2016).
+
+## Performance
+
+### Cityscapes
+
+| Model | Backbone | Resolution | Training Iters | mIoU | mIoU (flip) | mIoU (ms+flip) | Links |
+|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|
+|ENet|-|1024x512|80000|67.42%|68.11%|67.99%|[model](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/enet_cityscapes_1024x512_80k/model.pdparams)\|[log](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/enet_cityscapes_1024x512_80k/train.log)\|[vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=5d57386cdfcdb6a6bcb5135af134a0f2)|
--- a/configs/enet/enet_cityscapes_1024x512_80k.yml
+++ b/configs/enet/enet_cityscapes_1024x512_80k.yml
+_base_: '../_base_/cityscapes.yml'
+batch_size: 8
+
+train_dataset:
+  type: Cityscapes
+  dataset_root: data/cityscapes
+  transforms:
+    - type: RandomPaddingCrop
+      crop_size: [1024, 512]
+    - type: RandomDistort
+      brightness_range: 0.4
+      contrast_range: 0.4
+      saturation_range: 0.4
+    - type: Normalize
+  mode: train
+
+model:
+  type: ENet
+  num_classes: 19
+  pretrained: Null
+
+optimizer:
+  _inherited_: False
+  type: adam
+  weight_decay: 0.0002
+
+lr_scheduler:
+  end_lr: 0
+  learning_rate: 0.001
+  power: 0.9
+  type: PolynomialDecay
--- a/configs/espnet/README.md
+++ b/configs/espnet/README.md
+# ESPNetV2: A Light-weight, Power Efficient, and General Purpose Convolutional Neural Network
+
+## Reference
+
+> Mehta, Sachin, Mohammad Rastegari, Linda Shapiro, and Hannaneh Hajishirzi. "Espnetv2: A light-weight, power efficient, and general purpose convolutional neural network." In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 9190-9200. 2019.
+
+## Performance
+
+### CityScapes
+
+| Model | Backbone | Resolution | Training Iters | mIoU | Links |
+|:---:|:---:|:---:|:---:|:---:|:---:|
+|ESPNetV2|-|1024x512|120000|70.88%|[model](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/espnet_cityscapes_1024x512_120k/model.pdparams) \| [log](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/espnet_cityscapes_1024x512_120k/train.log) \|[vdl](https://www.paddlepaddle.org.cn/paddle/visualdl/service/app/scalar?id=c717bd8c2b5a083de759492158c14ffd)
+
+
+#### Additional Requirement
+- paddlepaddle develop version after 20211230
--- a/configs/espnet/espnet_cityscapes_1024x512_120k.yml
+++ b/configs/espnet/espnet_cityscapes_1024x512_120k.yml
+_base_: '../_base_/cityscapes.yml'
+
+batch_size: 8
+iters: 120000
+
+
+optimizer:
+  _inherited_: False
+  type: adam
+  weight_decay: 0.0002
+
+lr_scheduler:
+  type: PolynomialDecay
+  learning_rate: 0.001
+  end_lr: 0.0
+  power: 0.9
+
+loss:
+  types:
+    - type: CrossEntropyLoss
+      weight: [2.79834108 ,6.92945723 ,3.84068512 ,9.94349362 ,9.77098823 ,9.51484 ,10.30981624 ,9.94307377 ,4.64933892 ,9.55759938 ,7.86692178 ,9.53126629 ,10.3496365 ,6.67234062 ,10.26054204 ,10.28785275 ,10.28988296 ,10.40546021 ,10.13848367]
+  coef: [1, 1]
+
+model:
+  type: ESPNetV2
+  in_channels: 3
+  scale: 2.0
+  num_classes: 19
+  drop_prob: 0.0
--- a/configs/espnetv1/README.md
+++ b/configs/espnetv1/README.md
+# ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation
+
+
+## Reference
+
+> Mehta Sachin, Mohammad Rastegari, Anat Caspi, Linda Shapiro, and Hannaneh Hajishirzi. "ESPNet: Efficient Spatial Pyramid of Dilated Convolutions for Semantic Segmentation.".In Proceedings of the European Conference on Computer Vision, pp. 552-568. 2018.
+
+## Performance
+
+### Cityscapes
+
+| Model | Backbone | Resolution | Training Iters | mIoU | mIoU (flip) | mIoU (ms+flip) | Links |
+|:-:|:-:|:-:|:-:|:-:|:-:|:-:|:-:|
+|ESPNetV2|-|1024x512|120000|61.82%|62.20%|62.89%|[model](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/espnetv1_cityscapes_1024x512_120k/model.pdparams) \| [log](https://bj.bcebos.com/paddleseg/dygraph/cityscapes/espnetv1_cityscapes_1024x512_120k/train.log) \| [vdl](https://paddlepaddle.org.cn/paddle/visualdl/service/app?id=472e91a0600420c99a0dc3a1e6f80f87)