2021-08-01 00:31:58 +08:00
|
|
|
Collections:
|
2021-08-15 23:33:08 +08:00
|
|
|
- Metadata:
|
2021-08-01 00:31:58 +08:00
|
|
|
Training Data:
|
|
|
|
|
- Cityscapes
|
2021-08-15 23:33:08 +08:00
|
|
|
Name: fastscnn
|
2021-08-01 00:31:58 +08:00
|
|
|
Models:
|
2021-08-15 23:33:08 +08:00
|
|
|
- Config: configs/fastscnn/fast_scnn_lr0.12_8x4_160k_cityscapes.py
|
2021-08-01 00:31:58 +08:00
|
|
|
In Collection: fastscnn
|
|
|
|
|
Metadata:
|
|
|
|
|
backbone: Fast-SCNN
|
|
|
|
|
crop size: (512,1024)
|
|
|
|
|
inference time (ms/im):
|
2021-08-15 23:33:08 +08:00
|
|
|
- backend: PyTorch
|
2021-08-01 00:31:58 +08:00
|
|
|
batch size: 1
|
2021-08-15 23:33:08 +08:00
|
|
|
hardware: V100
|
2021-08-01 00:31:58 +08:00
|
|
|
mode: FP32
|
|
|
|
|
resolution: (512,1024)
|
2021-08-15 23:33:08 +08:00
|
|
|
value: 17.71
|
|
|
|
|
lr schd: 160000
|
2021-08-01 00:31:58 +08:00
|
|
|
memory (GB): 3.3
|
2021-08-15 23:33:08 +08:00
|
|
|
Name: fast_scnn_lr0.12_8x4_160k_cityscapes
|
2021-08-01 00:31:58 +08:00
|
|
|
Results:
|
|
|
|
|
Dataset: Cityscapes
|
|
|
|
|
Metrics:
|
|
|
|
|
mIoU: 70.96
|
|
|
|
|
mIoU(ms+flip): 72.65
|
2021-08-15 23:33:08 +08:00
|
|
|
Task: Semantic Segmentation
|
2021-08-09 18:13:09 +08:00
|
|
|
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/fast_scnn/fast_scnn_8x4_160k_lr0.12_cityscapes-0cec9937.pth
|
