Skip to content

Latest commit

 

History

History
35 lines (27 loc) · 1.09 KB

README.md

File metadata and controls

35 lines (27 loc) · 1.09 KB
Raw

Semantic Segmentation with VPD

Getting Started

  1. Install the mmsegmentation library and some required packages.
pip install openmim
mim install mmcv-full
mim install mmsegmentation
  1. Follow the guide in mmseg to prepare the ADE20k dataset.

Results and Fine-tuned Models

Model Config Head Crop Size Lr Schd mIoU mIoU (ms+flip) Fine-tuned Model
VPDSeg_SD-1-5 config Semantic FPN 512x512 80K 53.7 54.6 Tsinghua Cloud

Training

bash dist_train.sh <CONFIG_PATH> <NUM_GPUS>

We use 8 GPUs by default.

Evaluation

Command format:

bash dist_test.sh <CONFIG_PATH> <CHECKPOINT_PATH> <NUM_GPUS> --eval mIoU

To evaluate a model with multi-scale and flip, run

bash dist_test.sh <CONFIG_PATH> <CHECKPOINT_PATH> <NUM_GPUS> --eval mIoU --aug-test