Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -24,7 +24,7 @@ The first open-source, photorealistic synthetic dataset for single-shot fringe p
|
|
| 24 |
| Objects | 50 |
|
| 25 |
| Viewpoints per object | 6 (0°, 60°, 120°, 180°, 240°, 300°) |
|
| 26 |
| Resolution | 960 × 960 pixels |
|
| 27 |
-
| Measurement range | 1.5–1
|
| 28 |
| Ground truth method | 18-step phase shifting + Gray-code unwrapping |
|
| 29 |
| Train / Val / Test split | 240 / 30 / 30 (object-level, 40/5/5 objects) |
|
| 30 |
|
|
@@ -299,7 +299,7 @@ The split is performed at the **object level**—no object appears in more than
|
|
| 299 |
- **Synthetic only**: All data is rendered in simulation. Domain gap to real-world FPP systems has not been characterized. See the [paper](https://arxiv.org/abs/2601.08900) for discussion on sim-to-real transfer.
|
| 300 |
- **Material properties**: All objects use identical matte materials. Specular, translucent, or highly reflective surfaces are not represented.
|
| 301 |
- **Single-shot input**: Only `A_0.png` (first fringe image) is used as model input in the benchmarking study. The remaining 51 patterns are available for alternative formulations (e.g., multi-frame input).
|
| 302 |
-
- **Fixed measurement range**: All objects are scanned at 1.5–1
|
| 303 |
|
| 304 |
## Citation
|
| 305 |
|
|
|
|
| 24 |
| Objects | 50 |
|
| 25 |
| Viewpoints per object | 6 (0°, 60°, 120°, 180°, 240°, 300°) |
|
| 26 |
| Resolution | 960 × 960 pixels |
|
| 27 |
+
| Measurement range | 1.5–2.1 m |
|
| 28 |
| Ground truth method | 18-step phase shifting + Gray-code unwrapping |
|
| 29 |
| Train / Val / Test split | 240 / 30 / 30 (object-level, 40/5/5 objects) |
|
| 30 |
|
|
|
|
| 299 |
- **Synthetic only**: All data is rendered in simulation. Domain gap to real-world FPP systems has not been characterized. See the [paper](https://arxiv.org/abs/2601.08900) for discussion on sim-to-real transfer.
|
| 300 |
- **Material properties**: All objects use identical matte materials. Specular, translucent, or highly reflective surfaces are not represented.
|
| 301 |
- **Single-shot input**: Only `A_0.png` (first fringe image) is used as model input in the benchmarking study. The remaining 51 patterns are available for alternative formulations (e.g., multi-frame input).
|
| 302 |
+
- **Fixed measurement range**: All objects are scanned at 1.5–2.1 m. Performance at other distances is unknown.
|
| 303 |
|
| 304 |
## Citation
|
| 305 |
|