text stringclasses 9
values |
|---|
0 0.429688 0.326172 0.210938 0.591797 |
0 0.479980 0.476562 0.137695 0.322266 |
0 0.598633 0.371094 0.371094 0.679688 |
0 0.479492 0.120117 0.465820 0.134766 0.432617 0.145508 0.394531 0.140625 0.372070 0.123047 0.337891 0.159180 0.325195 0.195312 0.329102 0.280273 0.346680 0.458008 0.352539 0.485352 0.349609 0.497070 0.352539 0.540039 0.386719 0.606445 0.404297 0.618164 0.416992 0.621094 0.429688 0.619141 0.441406 0.603516 0.485352 0.5... |
1 0.375977 0.044922 0.369141 0.053711 0.367188 0.058594 0.368164 0.059570 0.367188 0.060547 0.367188 0.069336 0.368164 0.070312 0.367188 0.073242 0.368164 0.074219 0.368164 0.086914 0.369141 0.087891 0.369141 0.099609 0.370117 0.100586 0.369141 0.101562 0.371094 0.120117 0.383789 0.132812 0.392578 0.137695 0.404297 0.1... |
0 0.521484 0.336914 0.523438 0.355469 0.521484 0.366211 0.515625 0.375977 0.499023 0.386719 0.478516 0.390625 0.455078 0.385742 0.439453 0.373047 0.433594 0.354492 0.434570 0.338867 0.424805 0.356445 0.425781 0.384766 0.415039 0.399414 0.411133 0.412109 0.411133 0.444336 0.421875 0.601562 0.429688 0.615234 0.440430 0.6... |
1 0.451172 0.322266 0.445312 0.326172 0.436523 0.335938 0.437500 0.337891 0.436523 0.338867 0.435547 0.346680 0.434570 0.347656 0.435547 0.348633 0.434570 0.350586 0.435547 0.351562 0.435547 0.356445 0.436523 0.357422 0.437500 0.366211 0.440430 0.372070 0.444336 0.376953 0.445312 0.376953 0.449219 0.380859 0.457031 0.3... |
0 0.478516 0.235352 0.451172 0.254883 0.434570 0.274414 0.421875 0.297852 0.414062 0.329102 0.415039 0.406250 0.432617 0.625000 0.452148 0.666016 0.482422 0.692383 0.529297 0.708984 0.573242 0.708008 0.610352 0.694336 0.632812 0.677734 0.651367 0.653320 0.660156 0.631836 0.702148 0.375977 0.704102 0.342773 0.700195 0.3... |
1 0.779297 0.097656 0.714844 0.088867 0.644531 0.089844 0.634766 0.059570 0.617188 0.043945 0.587891 0.033203 0.561523 0.031250 0.543945 0.034180 0.516602 0.046875 0.502930 0.060547 0.494141 0.088867 0.496094 0.166992 0.486328 0.179688 0.480469 0.203125 0.484375 0.261719 0.492188 0.286133 0.510742 0.304688 0.533203 0.3... |
Synthetic Glass & Transparent Packaging Dataset Indoor (YOLO BBox + Segment)
A photorealistic synthetic computer vision dataset for transparent glass and packaging objects, designed for object detection and image segmentation under challenging lighting conditions. The dataset is intended for YOLO validation, transparent object detection, instance segmentation, synthetic-to-real experiments, and computer-vision research involving transparent and reflective materials.
This repository contains a 3-image sample dataset with dual ground-truth annotations so you can test pipeline compatibility instantly.
Dataset at a Glance
| Property | Details |
|---|---|
| Dataset type | Synthetic computer vision dataset |
| Objects | Transparent glass and packaging |
| Images | 240 |
| Resolution | 1024 × 1024 |
| Training split | 200 images |
| Validation split | 40 images |
| Detection annotations | YOLO bounding boxes |
| Segmentation annotations | YOLO instance segmentation polygons |
| Image format | PNG |
| Annotation format | YOLO .txt |
| Ground truth | Generated from 3D scene data |
| Primary tasks | Object detection and instance segmentation |
| Material challenges | Transparency, reflections, refractions, specular highlights |
| License | Commercial single-team license |
🛒 Download the Full Production Dataset (240 Images)
To fine-tune or evaluate your models on the complete dataset, access the full pack on Gumroad:
👉 Download Full Dataset on Gumroad ($49)
Full Dataset Specifications
240 Renders @ 1024x1024: 200 Train / 40 Val at native resolution.
Synthetic Computer Vision Dataset: Photorealistic renders of transparent glass and packaging objects.
Dual Ground-Truth Annotations: Includes both 2D Bounding Boxes (
labels_box/) and Multi-Point Instance Segmentation Polygons (labels_segment/).Physics-Based Ground Truth: Annotations are generated directly from the underlying 3D scene, avoiding manual labeling errors.
Challenging Transparent Materials: Designed to capture transparency, specular highlights, reflections, and surface refractions under difficult lighting conditions.
YOLO Compatible: Bounding boxes and instance segmentation annotations are provided in YOLO-compatible format.
Commercial Single-Team License: Royalty-free commercial usage rights to train, evaluate, and deploy models.
outdoor-pack : https://huggingface.co/datasets/Ji0134ch/Synthetic-Glass-Transparent-Packaging-Dataset-Sample
Repository Structure
.
├── data/
│ ├── images/ # Sample 1024x1024 PNG renders
│ ├── labels_box/ # YOLO format bounding box annotations (.txt)
│ └── labels_segment/ # YOLO format instance segmentation polygons (.txt)
├── src/
│ └──visualize_sample.py # Script to draw bounding box and polygon overlays
└── README.md
Applications
This synthetic glass object detection dataset can be used for:
- Transparent object detection
- Glass and transparent packaging detection
- YOLO object detection
- Instance segmentation
- Synthetic-to-real computer vision experiments
- Robotics and 3D vision research
- Downloads last month
- -