Skip to content
ModelListedExternal
Download

Manga109 Segmentation Bubble

A high-performance YOLO11n instance segmentation model fine-tuned for detecting and segmenting speech bubbles in manga/comic images.

KindOtherVersionRevision 6a098fa

Model source

Kind
Other
File size
11.4 MB
Version
vRevision 6a098fa
Updated
Sep 27, 2026
Downloads
46
Tasks
Image segmentation
Source
Hugging Face

Source excerpt

A high-performance YOLO11n instance segmentation model fine-tuned for detecting and segmenting speech bubbles in manga/comic images.

Sources

1 source
Hugging FacePrimaryVersion 6a098fa09edc393849f333acafa8230bbf6b3f2e
huggingface.co/juithealien/manga109-segmentation-bubble/tree/6a098fa09edc393849f333acafa8230bbf6b3f2e

Verified Sep 27

Model artifacts

2 artifacts

best.pt

pt · 11.4 MB · SHA-256 4028152940f7…0198 · Hugging Face

Download

last.pt

pt · 11.4 MB · SHA-256 9d672ebdd8b6…8275 · Hugging Face

Download

Source excerpts

2 excerpts

--- license: apache-2.0 language: - en - ja tags: - yolo - yolov11 - ultralytics - manga - speech-bubble - segmentation - object-detection - computer-vision - image-segmentation datasets: - MS92/MangaSegmentation - manga109 pipeline_tag: image-segmentation library_name: ultralytics --- # 🎯 MangaLens - Manga Speech Bubble Segmentation <div align="center"> ![Model](https://img.shields.io/badge/Model-YOLOv11n--seg-blue) ![Task](https://img.shields.io/badge/Task-Instance%20Segmentation-green) ![mAP50](https://img.shields.io/badge/mAP50-99.1%25-brightgreen) ![License](https://img.shields.io/badge/License-Apache%202.0-orange) </div> A high-performance YOLO11n instance segmentation model fine-tuned for detecting and segmenting speech bubbles in manga/comic images. ## �️ Demo Results <div align="center"> | Detection on Various Manga Styles | |:--:| | ![Demo 1](assets/demo1.jpg) | | *Speech bubble detection on action manga with multiple bubbles* | | ![Demo 2](assets/demo2.jpg) | |:--:| | *Detection on slice-of-life manga style* | </div> ## �📊 Model Performance ### Final Evaluation Results (Epoch 44) | Metric | Box Detection | Mask Segmentation | |--------|---------------|-------------------| | **Precision** | 97.55% | 97.66% | | **Recall** | 97.03% | 97.15% | | **mAP@50** | 99.10% | 99.13% | | **mAP@50-95** | 96.67% | 94.69% | ### Training Curves <div align="center"> ![Training Curves](assets/Code_Generated_Image.png) *Left: Segmentation Loss (Train vs Val) | Right: Mask mAP Metrics over epochs* </div> | Loss Type | Final Value | |-----------|-------------| | Box Loss | 0.2499 | | Segmentation Loss | 0.2762 | | Classification Loss | 0.2109 | | DFL Loss | 0.8064 | ## 🎓 Training Configuration | Parameter | Value | |-----------|-------| | Base Model | `yolo11n-seg.pt` | | Image Size | 1600×1600 | | Batch Size | 8 | | Epochs | 100 (Early stopped at 44) | | Optimizer | Auto (AdamW) | | Learning Rate | 0.01 | | Weight Decay | 0.0005 | | Patience | 10 | | AMP | Enabled | ### Data Augmentation - HSV Augmentation: H=0.015, S=0.7, V=0.4 - Mosaic: 1.0 - Flip Left-Right: 0.5 - Scale: 0.5 - Translate: 0.1 ## 📚 Training Data This model was trained on a combined dataset of: 1. **[MS92/MangaSegmentation](https://huggingface.co/datasets/MS92/MangaSegmentation)** - Manga panel and bubble segmentation dataset 2. **[Manga109](http://www.manga...

Source context: 46 downloads · 0 likes · Pipeline image-segmentation · Library ultralytics · Repo juithealien/manga109-segmentation-bubble