Skip to content
ModelVerifiedExternal
Download

Tbx11k Object Detection Faster Rcnn

This model is a Faster R-CNN detector fine-tuned on the TBX11K chest X-ray dataset for Tuberculosis localization and categorization.

Sourcehuggingface.co/nakasiga/tbx11k-object-detection-faster-rcnn/tree/649ed71ecc7c5c28ce977844a3d669f4ec0ebfbe
KindOtherVersionRevision 649ed71Publisher@nakasigaCgrade

Model source

Kind
Other
File size
330 MB
Version
vRevision 649ed71
Updated
Oct 21, 2026
Publisher
@nakasiga
Downloads
0
Source
Hugging Face

Source description

TBX11K Faster R-CNN (ResNet50-FPN-V2)

This model is a Faster R-CNN detector fine-tuned on the TBX11K chest X-ray dataset for Tuberculosis localization and categorization.

Training Configuration

  • Image Size: 512x512
  • Epochs: 30
  • Batch Size: 4
  • Optimization Strategy: WeightedRandomSampler for class balance control

Validation Performance (Best Epoch)

MetricScore
Best Validation [email protected]0.3728

Blind Test Inference Distribution

The model was run on 3302 unannotated test samples, outputting the following diagnosis breakdown:

  • Predicted Active TB: 420
  • Predicted Latent TB: 60
  • Predicted Healthy/Negative: 2822

Internal Model Class Map (Background Offset)

  • 0: Background (Internal Only)
  • 1: Active TB
  • 2: Latent TB

Sources

1 source
Hugging FacePrimaryVersion 649ed71ecc7c5c28ce977844a3d669f4ec0ebfbe
huggingface.co/nakasiga/tbx11k-object-detection-faster-rcnn/tree/649ed71ecc7c5c28ce977844a3d669f4ec0ebfbe

Verified Oct 21

Model artifacts

2 artifacts

best.pt

pt · 330 MB · SHA-256 814fca10d76b…364e · Hugging Face

Download

last.pt

pt · 330 MB · SHA-256 7835a4627392…bbc7 · Hugging Face

Download

Source excerpts

3 excerpts
ReadmeOct 21huggingface.co/nakasiga/tbx11k-object-detection-faster-rcnn

--- license: mit tags: - object-detection - medical - tuberculosis - faster-rcnn - chest-xray pipeline_tag: object-detection --- # TBX11K Faster R-CNN (ResNet50-FPN-V2) This model is a Faster R-CNN detector fine-tuned on the TBX11K chest X-ray dataset for Tuberculosis localization and categorization. ## Training Configuration - **Image Size**: 512x512 - **Epochs**: 30 - **Batch Size**: 4 - **Optimization Strategy**: WeightedRandomSampler for class balance control ## Validation Performance (Best Epoch) | Metric | Score | |--------|-------| | Best Validation [email protected] | 0.3728 | ## Blind Test Inference Distribution The model was run on 3302 unannotated test samples, outputting the following diagnosis breakdown: - **Predicted Active TB**: 420 - **Predicted Latent TB**: 60 - **Predicted Healthy/Negative**: 2822 ## Internal Model Class Map (Background Offset) - `0`: Background (Internal Only) - `1`: Active TB - `2`: Latent TB

ReadmeAug 5huggingface.co/nakasiga/tbx11k-object-detection-faster-rcnn

--- license: mit tags: - object-detection - medical - tuberculosis - faster-rcnn - chest-xray pipeline_tag: object-detection --- # TBX11K Faster R-CNN (ResNet50-FPN-V2) This model is a Faster R-CNN detector fine-tuned on the TBX11K chest X-ray dataset for Tuberculosis localization and categorization. ## Training Configuration - **Image Size**: 512x512 - **Epochs**: 30 - **Batch Size**: 4 - **Optimization Strategy**: WeightedRandomSampler for class balance control ## Validation Performance (Best Epoch) | Metric | Score | |--------|-------| | Best Validation [email protected] | 0.2217 | ## Blind Test Inference Distribution The model was run on 3302 unannotated test samples, outputting the following diagnosis breakdown: - **Predicted Active TB**: 468 - **Predicted Latent TB**: 40 - **Predicted Sick Non-TB**: 0 - **Predicted Healthy/Negative**: 2794 ## Internal Model Class Map (Background Offset) - `1`: Background (Internal Only) - `2`: Sick Non-TB - `3`: Active TB - `4`: Latent TB

ReadmeJul 13huggingface.co/nakasiga/tbx11k-object-detection-faster-rcnn

--- license: mit ---