tbx11k-object-detection-ssd

nakasiga/tbx11k-object-detection-ssd

tbx11k-object-detection-ssd is a object detection model from nakasiga. Use it when you need objects located in an image. The card lists the license as mit.

SSD300 (VGG16 backbone) optimized and trained on 11701 validated records from the TBX11K dataset.

Downloads · 30 days

Access

Public

Updated Jul 27, 2026

Repo size

384 MB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.pt192 MB · 100%

At a glance

Task: Object Detection
License: mit
Access: Public
Created: Jul 26, 2026
Updated: Jul 27, 2026
SHA: bafeebc5

Model card

From the Hugging Face model README

TBX11K SSD300 — Tuberculosis Object Detection

SSD300 (VGG16 backbone) optimized and trained on 11701 validated records from the TBX11K dataset.

Metric	Score
[email protected]	0.3892
[email protected]:0.95	0.2669
[email protected]	0.1446

Per-class Performance ([email protected] Threshold)

Class	[email protected]	Clinical Diagnostic Target
healthy	0.2669	Normal anatomical lung fields & latent signals
sick_non_tb	0.3372	Non-TB pulmonary anomalies (Pneumonia, Bronchitis)
active_tb	0.2987	Confirmed active Tuberculosis visual lesions

Global Summary Scores (900 Isolated Test Samples)

Global Test [email protected] : 0.3892 (38.9%)
Global Test COCO [email protected]:0.95 : 0.2669 (26.7%)
Global Test [email protected] : 0.1446 (14.5%)

Classes & Target Deployment Indices

0: __background__ (Strict PyTorch Detection Placeholder)
1: healthy (Includes clear lungs and latent tuberculosis signals)
2: sick_non_tb (Other pulmonary conditions / non-TB sickness)
3: active_tb (Active Tuberculosis visual lesions)

Model Configuration

Input Resolution: 300x300 pixels
Optimization Engine: AdamW + 5-Epoch Linear Warmup & Cosine Annealing
Oversampling Strategy: Square-Root Weighted Random Sampler targeting Active TB regions

Task: Object Detection
License: mit
Created: Jul 26, 2026
Updated: Jul 27, 2026