(6th) Solution for 2024 IEEE GRSS Data Fusion Contest Track 1
├── data
│ ├── dev
│ │ ├── p1
│ │ └── p2
│ └── Track1
│ ├── train
│ │ ├── images
│ │ └── labels
│ └── val
│ └── images
|
|
└── waterflow
|....
pip install -r requirements.txt
# infer
python run.py
Model = Unetpp + ResNeSt 269e
Loss = Diceloss + BCEloss -> rate=3:1
Pre-process = Normalize(img_mean, img_std)
Output = sigmoid + FindBest(threshold)
Trick = TTA
lr=3e-5, wd=4e-3, warm-up=0.2, ep=400
model\threshold | 0.5 | 0.3 | 0.1 |
---|---|---|---|
upp 269e | 0.9210 | ||
upp 269e + TTA | 0.9240 | 0.92560 | 0.9260 |
- K-Fold
- Visualize
- pre-process upgrade
- ...