Saltar al contenido
ModelPublicadoExterno
Descargar

Skin Lesion Segmentation Unet

This repository contains the pretrained model checkpoints accompanying the paper:

Fuentehuggingface.co/DevHabiba/skin-lesion-segmentation-unet/tree/6b117da14d23f95680bf694fedd7de1a1734a58e
TipoOtherVersiónRevision 6b117daDcalidad

Fuente del modelo

Tipo
Other
Tamaño del archivo
182–547 MB
Versión
vRevision 6b117da
Actualizado
5 oct 2026
Descargas
0
Fuente
Hugging Face

Descripción de la fuente

Teacher–Student Pseudo-Labeling for Semi-Supervised Skin Lesion Segmentation

This repository contains the pretrained model checkpoints accompanying the paper:

Teacher–Student Pseudo-Labeling for Semi-Supervised Skin Lesion Segmentation: A Reproducible Baseline

The models implement a teacher–student pseudo-labeling framework for semantic segmentation of dermoscopic skin lesions using a U-Net architecture with a pretrained ResNet-34 encoder.

Overview

Pixel-wise annotation of medical images is expensive and requires expert dermatologists. This work investigates a simple yet effective semi-supervised learning strategy that leverages unlabeled dermoscopic images through teacher-generated pseudo labels.

The repository provides pretrained checkpoints for:

  • Fully supervised baseline
  • Teacher model (20% labeled data)
  • Student model (20% labeled + pseudo labels)
  • Teacher model (50% labeled data)
  • Student model (50% labeled + pseudo labels)

The implementation is fully reproducible and was trained using free Kaggle Tesla T4 GPU resources.

Model Architecture

  • Architecture: U-Net
  • Encoder: ResNet-34 (ImageNet pretrained)
  • Framework: PyTorch
  • Image Size: 256 × 256
  • Task: Binary semantic segmentation
  • Dataset: ISIC 2018 Skin Lesion Segmentation Challenge

Available Checkpoints

ModelDescription
best_full_supervised_model.pthFully supervised baseline trained using 100% labeled images
teacher_20pct.pthTeacher model trained using only 20% labeled images
student_20pct.pthStudent model trained using 20% labeled images and pseudo labels
teacher_50pct.pthTeacher model trained using only 50% labeled images
student_50pct.pthStudent model trained using 50% labeled images and pseudo labels

Performance

Fully Supervised Baseline

MetricScore
Dice0.7769
IoU0.6482
Pixel Accuracy0.9130

Semi-Supervised Results

ExperimentBest Validation Dice
Teacher (20%)0.7570
Student (20%)0.7638
Teacher (50%)0.7759
Student (50%)0.7800

Using only 20% labeled data, the student model recovers 98.3% of the fully supervised Dice score.

Using 50% labeled data, the student slightly surpasses the fully supervised baseline.

Ablation Study

Confidence threshold analysis for pseudo-label selection:

ThresholdAccepted MasksDice
0.6017430.7506
0.7016850.7551
0.8015630.7638
0.9011540.7604

The optimal threshold was τ = 0.80, balancing pseudo-label quality and quantity.

Loading a Checkpoint

import torch

model = YourUNetModel()

checkpoint = torch.load("student_50pct.pth", map_location="cpu")

model.load_state_dict(checkpoint)

model.eval()

Intended Use

These checkpoints are intended for:

  • Medical image segmentation research
  • Semi-supervised learning research
  • Benchmark comparisons
  • Educational purposes
  • Reproducibility of the accompanying paper

They are not intended for clinical diagnosis or direct patient care.

Related Resources

Citation

If you use this work in your research, please cite:

@misc{sajid2026teacher,
  author       = {Habiba Sajid},
  title        = {Teacher--Student Pseudo-Labeling for Semi-Supervised Skin Lesion Segmentation: A Reproducible Baseline},
  year         = {2026},
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.21320667},
  url          = {https://doi.org/10.5281/zenodo.21320667}
}

---

# License

Released under the MIT License.

Fuentes

1 fuente
Hugging FacePrincipalVersión 6b117da14d23f95680bf694fedd7de1a1734a58e
huggingface.co/DevHabiba/skin-lesion-segmentation-unet/tree/6b117da14d23f95680bf694fedd7de1a1734a58e

Verificado 3 sept

Artefactos del modelo

5 artefactos

best_full_supervised_model.pth

pth · 547 MB · SHA-256 2b3f0f1dfab5…e3f7 · Hugging Face

Descargar

best_student_semisup_20pct.pth

pth · 182 MB · SHA-256 02c664cabaa6…a96f · Hugging Face

Descargar

best_student_semisup_50pct.pth

pth · 182 MB · SHA-256 943a9a3816bf…0125 · Hugging Face

Descargar

best_teacher_semisup_20pct.pth

pth · 182 MB · SHA-256 61d737ecec90…9d47 · Hugging Face

Descargar

best_teacher_semisup_50pct.pth

pth · 182 MB · SHA-256 833de94eb23e…3ba8 · Hugging Face

Descargar

Extractos de fuentes

3 extractos
Readme16 julhuggingface.co/DevHabiba/skin-lesion-segmentation-unet

# Teacher–Student Pseudo-Labeling for Semi-Supervised Skin Lesion Segmentation This repository contains the pretrained model checkpoints accompanying the paper: > **Teacher–Student Pseudo-Labeling for Semi-Supervised Skin Lesion Segmentation: A Reproducible Baseline** The models implement a teacher–student pseudo-labeling framework for semantic segmentation of dermoscopic skin lesions using a U-Net architecture with a pretrained ResNet-34 encoder. --- # Overview Pixel-wise annotation of medical images is expensive and requires expert dermatologists. This work investigates a simple yet effective semi-supervised learning strategy that leverages unlabeled dermoscopic images through teacher-generated pseudo labels. The repository provides pretrained checkpoints for: - Fully supervised baseline - Teacher model (20% labeled data) - Student model (20% labeled + pseudo labels) - Teacher model (50% labeled data) - Student model (50% labeled + pseudo labels) The implementation is fully reproducible and was trained using free Kaggle Tesla T4 GPU resources. --- # Model Architecture - Architecture: U-Net - Encoder: ResNet-34 (ImageNet pretrained) - Framework: PyTorch - Image Size: 256 × 256 - Task: Binary semantic segmentation - Dataset: ISIC 2018 Skin Lesion Segmentation Challenge --- # Available Checkpoints | Model | Description | |--------|-------------| | `best_full_supervised_model.pth` | Fully supervised baseline trained using 100% labeled images | | `teacher_20pct.pth` | Teacher model trained using only 20% labeled images | | `student_20pct.pth` | Student model trained using 20% labeled images and pseudo labels | | `teacher_50pct.pth` | Teacher model trained using only 50% labeled images | | `student_50pct.pth` | Student model trained using 50% labeled images and pseudo labels | --- # Performance ## Fully Supervised Baseline | Metric | Score | |--------|------:| | Dice | **0.7769** | | IoU | 0.6482 | | Pixel Accuracy | 0.9130 | --- ## Semi-Supervised Results | Experiment | Best Validation Dice | |------------|--------------------:| | Teacher (20%) | 0.7570 | | Student (20%) | **0.7638** | | Teacher (50%) | 0.7759 | | Student (50%) | **0.7800** | Using only **20% labeled data**, the student model recovers **98.3%** of the fully supervised Dice score. Using **50% labeled data**, the student slightly surpasses the fully supervised baseline....

Readme11 julhuggingface.co/DevHabiba/skin-lesion-segmentation-unet

--- license: mit language: - en tags: - medical-image-segmentation - skin-lesion-segmentation - semantic-segmentation - semi-supervised-learning - pseudo-labeling - pytorch - resnet34 - unet --- dataset: - ISIC 2018 # Teacher–Student Pseudo-Labeling for Semi-Supervised Skin Lesion Segmentation This repository contains the pretrained model checkpoints accompanying the paper: > **Teacher–Student Pseudo-Labeling for Semi-Supervised Skin Lesion Segmentation: A Reproducible Baseline** The models implement a teacher–student pseudo-labeling framework for semantic segmentation of dermoscopic skin lesions using a U-Net architecture with a pretrained ResNet-34 encoder. --- # Overview Pixel-wise annotation of medical images is expensive and requires expert dermatologists. This work investigates a simple yet effective semi-supervised learning strategy that leverages unlabeled dermoscopic images through teacher-generated pseudo labels. The repository provides pretrained checkpoints for: - Fully supervised baseline - Teacher model (20% labeled data) - Student model (20% labeled + pseudo labels) - Teacher model (50% labeled data) - Student model (50% labeled + pseudo labels) The implementation is fully reproducible and was trained using free Kaggle Tesla T4 GPU resources. --- # Model Architecture - Architecture: U-Net - Encoder: ResNet-34 (ImageNet pretrained) - Framework: PyTorch - Image Size: 256 × 256 - Task: Binary semantic segmentation - Dataset: ISIC 2018 Skin Lesion Segmentation Challenge --- # Available Checkpoints | Model | Description | |--------|-------------| | `best_full_supervised_model.pth` | Fully supervised baseline trained using 100% labeled images | | `teacher_20pct.pth` | Teacher model trained using only 20% labeled images | | `student_20pct.pth` | Student model trained using 20% labeled images and pseudo labels | | `teacher_50pct.pth` | Teacher model trained using only 50% labeled images | | `student_50pct.pth` | Student model trained using 50% labeled images and pseudo labels | --- # Performance ## Fully Supervised Baseline | Metric | Score | |--------|------:| | Dice | **0.7769** | | IoU | 0.6482 | | Pixel Accuracy | 0.9130 | --- ## Semi-Supervised Results | Experiment | Best Validation Dice | |------------|--------------------:| | Teacher (20%) | 0.7570 | | Student (20%) | **0.7638** | | Teacher (50%) | 0.7759 | | Student (5...

Note11 julhuggingface.co/DevHabiba/skin-lesion-segmentation-unet

Source context: 0 downloads · 0 likes · Repo DevHabiba/skin-lesion-segmentation-unet