README.md

---
base_model: DCAMA
language: en
license: mit
tags:
- few-shot segmentation
- distillation
- image-segmentation
name: DistillFSS-DCAMA
library: pytorch
ArXiv: '2512.05613'
repo_url: https://github.com/pasqualedem/DistillFSS
paper_url: https://arxiv.org/abs/2512.05613
parameters: "dataloader:\n  num_workers: 0\ndataset:\n  datasets:\n    test_weedmap:\n\
  \      prompt_images: 5\n      test_root: data/weedmap/0_rotations_processed_003_test/RedEdge/003\n\
  \      train_root: data/weedmap/0_rotations_processed_003_test/RedEdge/000\n  preprocess:\n\
  \    image_size: 384\n    mean:\n    - 0.485\n    - 0.456\n    - 0.406\n    std:\n\
  \    - 0.229\n    - 0.224\n    - 0.225\nmodel:\n  name: distillator\n  params:\n\
  \    student:\n      name: conv_distillator\n      num_classes: 2\n    teacher:\n\
  \      backbone: swin\n      backbone_checkpoint: checkpoints/swin_base_patch4_window12_384.pth\n\
  \      concat_support: false\n      image_size: 384\n      model_checkpoint: checkpoints/swin_fold0_pascal_modcross_soft.pt\n\
  \      name: dcama\npush_to_hub:\n  repo_name: pasqualedem/DistillFSS_WeedMap_DCAMA_5shot\n\
  refinement:\n  hot_parameters:\n  - model.conv1\n  - model.conv2\n  - model.conv3\n\
  \  - model.mixer1\n  - model.mixer2\n  - model.mixer3\n  - student\n  iterations_is_num_classes:\
  \ false\n  loss:\n    name: refine_distill\n  lr: 0.001\n  max_iterations: 500\n\
  \  subsample: 1\n  substitutor: paired\ntest:\n  prompt_to_use: null\ntracker:\n\
  \  cache_dir: tmp\n  group: WeedMap\n  log_frequency: 1\n  project: FSSWeed\n  tags:\n\
  \  - WeedMap\n  - Distill\n  test_image_log_frequency: 10\n  tmp_dir: tmp\n  train_image_log_frequency:\
  \ 25\n"
repo_id: pasqualedem/DistillFSS_WeedMap_DCAMA_5shot
---

DistillFSS-DCAMA is a distilled version of the DCAMA model for a specific downstream segmentation task. The DistillFSS framework allows to distill large few-shot segmentation models into smaller and more efficient ones, while improving or maintaining their performance on the target task. 

- Code: https://github.com/pasqualedem/DistillFSS
- Paper: https://arxiv.org/abs/2512.05613

How to use this model:
Clone the repository: 
```bash
git clone https://github.com/pasqualedem/DistillFSS.git
```
Install the required dependencies as specified in the repository.

Load the model using the following code snippet:
```python
from distillfss.models.dcama.distillator import DistilledDCAMA
model = DistilledDCAMA.from_pretrained("pasqualedem/DistillFSS_WeedMap_DCAMA_5shot")
```

YAML configuration:
```yaml
dataloader:
  num_workers: 0
dataset:
  datasets:
    test_weedmap:
      prompt_images: 5
      test_root: data/weedmap/0_rotations_processed_003_test/RedEdge/003
      train_root: data/weedmap/0_rotations_processed_003_test/RedEdge/000
  preprocess:
    image_size: 384
    mean:
    - 0.485
    - 0.456
    - 0.406
    std:
    - 0.229
    - 0.224
    - 0.225
model:
  name: distillator
  params:
    student:
      name: conv_distillator
      num_classes: 2
    teacher:
      backbone: swin
      backbone_checkpoint: checkpoints/swin_base_patch4_window12_384.pth
      concat_support: false
      image_size: 384
      model_checkpoint: checkpoints/swin_fold0_pascal_modcross_soft.pt
      name: dcama
push_to_hub:
  repo_name: pasqualedem/DistillFSS_WeedMap_DCAMA_5shot
refinement:
  hot_parameters:
  - model.conv1
  - model.conv2
  - model.conv3
  - model.mixer1
  - model.mixer2
  - model.mixer3
  - student
  iterations_is_num_classes: false
  loss:
    name: refine_distill
  lr: 0.001
  max_iterations: 500
  subsample: 1
  substitutor: paired
test:
  prompt_to_use: null
tracker:
  cache_dir: tmp
  group: WeedMap
  log_frequency: 1
  project: FSSWeed
  tags:
  - WeedMap
  - Distill
  test_image_log_frequency: 10
  tmp_dir: tmp
  train_image_log_frequency: 25

```