Datasets
Models
Downloads: 1
Diff of
/configs/ccnet/ccnet.yml
[000000]
..
[4e96d3]
Switch to unified view
| a | b/configs/ccnet/ccnet.yml | ||
|---|---|---|---|
| 1 | Collections: |
||
| 2 | - Name: ccnet |
||
| 3 | Metadata: |
||
| 4 | Training Data: |
||
| 5 | - Cityscapes |
||
| 6 | - ADE20K |
||
| 7 | - Pascal VOC 2012 + Aug |
||
| 8 | Paper: |
||
| 9 | URL: https://arxiv.org/abs/1811.11721 |
||
| 10 | Title: 'CCNet: Criss-Cross Attention for Semantic Segmentation' |
||
| 11 | README: configs/ccnet/README.md |
||
| 12 | Code: |
||
| 13 | URL: https://github.com/open-mmlab/mmsegmentation/blob/v0.17.0/mmseg/models/decode_heads/apc_head.py#L111 |
||
| 14 | Version: v0.17.0 |
||
| 15 | Converted From: |
||
| 16 | Code: https://github.com/speedinghzl/CCNet |
||
| 17 | Models: |
||
| 18 | - Name: ccnet_r50-d8_512x1024_40k_cityscapes |
||
| 19 | In Collection: ccnet |
||
| 20 | Metadata: |
||
| 21 | backbone: R-50-D8 |
||
| 22 | crop size: (512,1024) |
||
| 23 | lr schd: 40000 |
||
| 24 | inference time (ms/im): |
||
| 25 | - value: 301.2 |
||
| 26 | hardware: V100 |
||
| 27 | backend: PyTorch |
||
| 28 | batch size: 1 |
||
| 29 | mode: FP32 |
||
| 30 | resolution: (512,1024) |
||
| 31 | Training Memory (GB): 6.0 |
||
| 32 | Results: |
||
| 33 | - Task: Semantic Segmentation |
||
| 34 | Dataset: Cityscapes |
||
| 35 | Metrics: |
||
| 36 | mIoU: 77.76 |
||
| 37 | mIoU(ms+flip): 78.87 |
||
| 38 | Config: configs/ccnet/ccnet_r50-d8_512x1024_40k_cityscapes.py |
||
| 39 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x1024_40k_cityscapes/ccnet_r50-d8_512x1024_40k_cityscapes_20200616_142517-4123f401.pth |
||
| 40 | - Name: ccnet_r101-d8_512x1024_40k_cityscapes |
||
| 41 | In Collection: ccnet |
||
| 42 | Metadata: |
||
| 43 | backbone: R-101-D8 |
||
| 44 | crop size: (512,1024) |
||
| 45 | lr schd: 40000 |
||
| 46 | inference time (ms/im): |
||
| 47 | - value: 432.9 |
||
| 48 | hardware: V100 |
||
| 49 | backend: PyTorch |
||
| 50 | batch size: 1 |
||
| 51 | mode: FP32 |
||
| 52 | resolution: (512,1024) |
||
| 53 | Training Memory (GB): 9.5 |
||
| 54 | Results: |
||
| 55 | - Task: Semantic Segmentation |
||
| 56 | Dataset: Cityscapes |
||
| 57 | Metrics: |
||
| 58 | mIoU: 76.35 |
||
| 59 | mIoU(ms+flip): 78.19 |
||
| 60 | Config: configs/ccnet/ccnet_r101-d8_512x1024_40k_cityscapes.py |
||
| 61 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_512x1024_40k_cityscapes/ccnet_r101-d8_512x1024_40k_cityscapes_20200616_142540-a3b84ba6.pth |
||
| 62 | - Name: ccnet_r50-d8_769x769_40k_cityscapes |
||
| 63 | In Collection: ccnet |
||
| 64 | Metadata: |
||
| 65 | backbone: R-50-D8 |
||
| 66 | crop size: (769,769) |
||
| 67 | lr schd: 40000 |
||
| 68 | inference time (ms/im): |
||
| 69 | - value: 699.3 |
||
| 70 | hardware: V100 |
||
| 71 | backend: PyTorch |
||
| 72 | batch size: 1 |
||
| 73 | mode: FP32 |
||
| 74 | resolution: (769,769) |
||
| 75 | Training Memory (GB): 6.8 |
||
| 76 | Results: |
||
| 77 | - Task: Semantic Segmentation |
||
| 78 | Dataset: Cityscapes |
||
| 79 | Metrics: |
||
| 80 | mIoU: 78.46 |
||
| 81 | mIoU(ms+flip): 79.93 |
||
| 82 | Config: configs/ccnet/ccnet_r50-d8_769x769_40k_cityscapes.py |
||
| 83 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_769x769_40k_cityscapes/ccnet_r50-d8_769x769_40k_cityscapes_20200616_145125-76d11884.pth |
||
| 84 | - Name: ccnet_r101-d8_769x769_40k_cityscapes |
||
| 85 | In Collection: ccnet |
||
| 86 | Metadata: |
||
| 87 | backbone: R-101-D8 |
||
| 88 | crop size: (769,769) |
||
| 89 | lr schd: 40000 |
||
| 90 | inference time (ms/im): |
||
| 91 | - value: 990.1 |
||
| 92 | hardware: V100 |
||
| 93 | backend: PyTorch |
||
| 94 | batch size: 1 |
||
| 95 | mode: FP32 |
||
| 96 | resolution: (769,769) |
||
| 97 | Training Memory (GB): 10.7 |
||
| 98 | Results: |
||
| 99 | - Task: Semantic Segmentation |
||
| 100 | Dataset: Cityscapes |
||
| 101 | Metrics: |
||
| 102 | mIoU: 76.94 |
||
| 103 | mIoU(ms+flip): 78.62 |
||
| 104 | Config: configs/ccnet/ccnet_r101-d8_769x769_40k_cityscapes.py |
||
| 105 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_769x769_40k_cityscapes/ccnet_r101-d8_769x769_40k_cityscapes_20200617_101428-4f57c8d0.pth |
||
| 106 | - Name: ccnet_r50-d8_512x1024_80k_cityscapes |
||
| 107 | In Collection: ccnet |
||
| 108 | Metadata: |
||
| 109 | backbone: R-50-D8 |
||
| 110 | crop size: (512,1024) |
||
| 111 | lr schd: 80000 |
||
| 112 | Results: |
||
| 113 | - Task: Semantic Segmentation |
||
| 114 | Dataset: Cityscapes |
||
| 115 | Metrics: |
||
| 116 | mIoU: 79.03 |
||
| 117 | mIoU(ms+flip): 80.16 |
||
| 118 | Config: configs/ccnet/ccnet_r50-d8_512x1024_80k_cityscapes.py |
||
| 119 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x1024_80k_cityscapes/ccnet_r50-d8_512x1024_80k_cityscapes_20200617_010421-869a3423.pth |
||
| 120 | - Name: ccnet_r101-d8_512x1024_80k_cityscapes |
||
| 121 | In Collection: ccnet |
||
| 122 | Metadata: |
||
| 123 | backbone: R-101-D8 |
||
| 124 | crop size: (512,1024) |
||
| 125 | lr schd: 80000 |
||
| 126 | Results: |
||
| 127 | - Task: Semantic Segmentation |
||
| 128 | Dataset: Cityscapes |
||
| 129 | Metrics: |
||
| 130 | mIoU: 78.87 |
||
| 131 | mIoU(ms+flip): 79.9 |
||
| 132 | Config: configs/ccnet/ccnet_r101-d8_512x1024_80k_cityscapes.py |
||
| 133 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_512x1024_80k_cityscapes/ccnet_r101-d8_512x1024_80k_cityscapes_20200617_203935-ffae8917.pth |
||
| 134 | - Name: ccnet_r50-d8_769x769_80k_cityscapes |
||
| 135 | In Collection: ccnet |
||
| 136 | Metadata: |
||
| 137 | backbone: R-50-D8 |
||
| 138 | crop size: (769,769) |
||
| 139 | lr schd: 80000 |
||
| 140 | Results: |
||
| 141 | - Task: Semantic Segmentation |
||
| 142 | Dataset: Cityscapes |
||
| 143 | Metrics: |
||
| 144 | mIoU: 79.29 |
||
| 145 | mIoU(ms+flip): 81.08 |
||
| 146 | Config: configs/ccnet/ccnet_r50-d8_769x769_80k_cityscapes.py |
||
| 147 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_769x769_80k_cityscapes/ccnet_r50-d8_769x769_80k_cityscapes_20200617_010421-73eed8ca.pth |
||
| 148 | - Name: ccnet_r101-d8_769x769_80k_cityscapes |
||
| 149 | In Collection: ccnet |
||
| 150 | Metadata: |
||
| 151 | backbone: R-101-D8 |
||
| 152 | crop size: (769,769) |
||
| 153 | lr schd: 80000 |
||
| 154 | Results: |
||
| 155 | - Task: Semantic Segmentation |
||
| 156 | Dataset: Cityscapes |
||
| 157 | Metrics: |
||
| 158 | mIoU: 79.45 |
||
| 159 | mIoU(ms+flip): 80.66 |
||
| 160 | Config: configs/ccnet/ccnet_r101-d8_769x769_80k_cityscapes.py |
||
| 161 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_769x769_80k_cityscapes/ccnet_r101-d8_769x769_80k_cityscapes_20200618_011502-ad3cd481.pth |
||
| 162 | - Name: ccnet_r50-d8_512x512_80k_ade20k |
||
| 163 | In Collection: ccnet |
||
| 164 | Metadata: |
||
| 165 | backbone: R-50-D8 |
||
| 166 | crop size: (512,512) |
||
| 167 | lr schd: 80000 |
||
| 168 | inference time (ms/im): |
||
| 169 | - value: 47.87 |
||
| 170 | hardware: V100 |
||
| 171 | backend: PyTorch |
||
| 172 | batch size: 1 |
||
| 173 | mode: FP32 |
||
| 174 | resolution: (512,512) |
||
| 175 | Training Memory (GB): 8.8 |
||
| 176 | Results: |
||
| 177 | - Task: Semantic Segmentation |
||
| 178 | Dataset: ADE20K |
||
| 179 | Metrics: |
||
| 180 | mIoU: 41.78 |
||
| 181 | mIoU(ms+flip): 42.98 |
||
| 182 | Config: configs/ccnet/ccnet_r50-d8_512x512_80k_ade20k.py |
||
| 183 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x512_80k_ade20k/ccnet_r50-d8_512x512_80k_ade20k_20200615_014848-aa37f61e.pth |
||
| 184 | - Name: ccnet_r101-d8_512x512_80k_ade20k |
||
| 185 | In Collection: ccnet |
||
| 186 | Metadata: |
||
| 187 | backbone: R-101-D8 |
||
| 188 | crop size: (512,512) |
||
| 189 | lr schd: 80000 |
||
| 190 | inference time (ms/im): |
||
| 191 | - value: 70.87 |
||
| 192 | hardware: V100 |
||
| 193 | backend: PyTorch |
||
| 194 | batch size: 1 |
||
| 195 | mode: FP32 |
||
| 196 | resolution: (512,512) |
||
| 197 | Training Memory (GB): 12.2 |
||
| 198 | Results: |
||
| 199 | - Task: Semantic Segmentation |
||
| 200 | Dataset: ADE20K |
||
| 201 | Metrics: |
||
| 202 | mIoU: 43.97 |
||
| 203 | mIoU(ms+flip): 45.13 |
||
| 204 | Config: configs/ccnet/ccnet_r101-d8_512x512_80k_ade20k.py |
||
| 205 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_512x512_80k_ade20k/ccnet_r101-d8_512x512_80k_ade20k_20200615_014848-1f4929a3.pth |
||
| 206 | - Name: ccnet_r50-d8_512x512_160k_ade20k |
||
| 207 | In Collection: ccnet |
||
| 208 | Metadata: |
||
| 209 | backbone: R-50-D8 |
||
| 210 | crop size: (512,512) |
||
| 211 | lr schd: 160000 |
||
| 212 | Results: |
||
| 213 | - Task: Semantic Segmentation |
||
| 214 | Dataset: ADE20K |
||
| 215 | Metrics: |
||
| 216 | mIoU: 42.08 |
||
| 217 | mIoU(ms+flip): 43.13 |
||
| 218 | Config: configs/ccnet/ccnet_r50-d8_512x512_160k_ade20k.py |
||
| 219 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x512_160k_ade20k/ccnet_r50-d8_512x512_160k_ade20k_20200616_084435-7c97193b.pth |
||
| 220 | - Name: ccnet_r101-d8_512x512_160k_ade20k |
||
| 221 | In Collection: ccnet |
||
| 222 | Metadata: |
||
| 223 | backbone: R-101-D8 |
||
| 224 | crop size: (512,512) |
||
| 225 | lr schd: 160000 |
||
| 226 | Results: |
||
| 227 | - Task: Semantic Segmentation |
||
| 228 | Dataset: ADE20K |
||
| 229 | Metrics: |
||
| 230 | mIoU: 43.71 |
||
| 231 | mIoU(ms+flip): 45.04 |
||
| 232 | Config: configs/ccnet/ccnet_r101-d8_512x512_160k_ade20k.py |
||
| 233 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_512x512_160k_ade20k/ccnet_r101-d8_512x512_160k_ade20k_20200616_000644-e849e007.pth |
||
| 234 | - Name: ccnet_r50-d8_512x512_20k_voc12aug |
||
| 235 | In Collection: ccnet |
||
| 236 | Metadata: |
||
| 237 | backbone: R-50-D8 |
||
| 238 | crop size: (512,512) |
||
| 239 | lr schd: 20000 |
||
| 240 | inference time (ms/im): |
||
| 241 | - value: 48.9 |
||
| 242 | hardware: V100 |
||
| 243 | backend: PyTorch |
||
| 244 | batch size: 1 |
||
| 245 | mode: FP32 |
||
| 246 | resolution: (512,512) |
||
| 247 | Training Memory (GB): 6.0 |
||
| 248 | Results: |
||
| 249 | - Task: Semantic Segmentation |
||
| 250 | Dataset: Pascal VOC 2012 + Aug |
||
| 251 | Metrics: |
||
| 252 | mIoU: 76.17 |
||
| 253 | mIoU(ms+flip): 77.51 |
||
| 254 | Config: configs/ccnet/ccnet_r50-d8_512x512_20k_voc12aug.py |
||
| 255 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x512_20k_voc12aug/ccnet_r50-d8_512x512_20k_voc12aug_20200617_193212-fad81784.pth |
||
| 256 | - Name: ccnet_r101-d8_512x512_20k_voc12aug |
||
| 257 | In Collection: ccnet |
||
| 258 | Metadata: |
||
| 259 | backbone: R-101-D8 |
||
| 260 | crop size: (512,512) |
||
| 261 | lr schd: 20000 |
||
| 262 | inference time (ms/im): |
||
| 263 | - value: 73.31 |
||
| 264 | hardware: V100 |
||
| 265 | backend: PyTorch |
||
| 266 | batch size: 1 |
||
| 267 | mode: FP32 |
||
| 268 | resolution: (512,512) |
||
| 269 | Training Memory (GB): 9.5 |
||
| 270 | Results: |
||
| 271 | - Task: Semantic Segmentation |
||
| 272 | Dataset: Pascal VOC 2012 + Aug |
||
| 273 | Metrics: |
||
| 274 | mIoU: 77.27 |
||
| 275 | mIoU(ms+flip): 79.02 |
||
| 276 | Config: configs/ccnet/ccnet_r101-d8_512x512_20k_voc12aug.py |
||
| 277 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_512x512_20k_voc12aug/ccnet_r101-d8_512x512_20k_voc12aug_20200617_193212-0007b61d.pth |
||
| 278 | - Name: ccnet_r50-d8_512x512_40k_voc12aug |
||
| 279 | In Collection: ccnet |
||
| 280 | Metadata: |
||
| 281 | backbone: R-50-D8 |
||
| 282 | crop size: (512,512) |
||
| 283 | lr schd: 40000 |
||
| 284 | Results: |
||
| 285 | - Task: Semantic Segmentation |
||
| 286 | Dataset: Pascal VOC 2012 + Aug |
||
| 287 | Metrics: |
||
| 288 | mIoU: 75.96 |
||
| 289 | mIoU(ms+flip): 77.04 |
||
| 290 | Config: configs/ccnet/ccnet_r50-d8_512x512_40k_voc12aug.py |
||
| 291 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x512_40k_voc12aug/ccnet_r50-d8_512x512_40k_voc12aug_20200613_232127-c2a15f02.pth |
||
| 292 | - Name: ccnet_r101-d8_512x512_40k_voc12aug |
||
| 293 | In Collection: ccnet |
||
| 294 | Metadata: |
||
| 295 | backbone: R-101-D8 |
||
| 296 | crop size: (512,512) |
||
| 297 | lr schd: 40000 |
||
| 298 | Results: |
||
| 299 | - Task: Semantic Segmentation |
||
| 300 | Dataset: Pascal VOC 2012 + Aug |
||
| 301 | Metrics: |
||
| 302 | mIoU: 77.87 |
||
| 303 | mIoU(ms+flip): 78.9 |
||
| 304 | Config: configs/ccnet/ccnet_r101-d8_512x512_40k_voc12aug.py |
||
| 305 | Weights: https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_512x512_40k_voc12aug/ccnet_r101-d8_512x512_40k_voc12aug_20200613_232127-c30da577.pth |