Datasets:
text stringlengths 37 37 |
|---|
0 0.564844 0.523438 0.307812 0.418750 |
0 0.337500 0.381250 0.018750 0.062500 |
0 0.494531 0.531250 0.098437 0.390625 |
0 0.335938 0.385156 0.018750 0.054688 |
0 0.432812 0.568750 0.193750 0.300000 |
0 0.346875 0.380469 0.021875 0.051562 |
0 0.435156 0.583594 0.195312 0.295312 |
0 0.364844 0.390625 0.032813 0.040625 |
0 0.437500 0.582812 0.190625 0.303125 |
0 0.367188 0.384375 0.025000 0.056250 |
0 0.437500 0.585938 0.193750 0.300000 |
0 0.367969 0.386719 0.035937 0.051562 |
0 0.437500 0.590625 0.196875 0.290625 |
0 0.364063 0.391406 0.034375 0.042188 |
0 0.437500 0.590625 0.196875 0.293750 |
0 0.370312 0.383594 0.025000 0.057813 |
0 0.439844 0.587500 0.195312 0.300000 |
0 0.368750 0.388281 0.037500 0.048438 |
0 0.439063 0.588281 0.196875 0.298438 |
0 0.366406 0.390625 0.032813 0.043750 |
0 0.439063 0.588281 0.193750 0.298438 |
0 0.369531 0.384375 0.023438 0.056250 |
0 0.438281 0.591406 0.195312 0.292187 |
0 0.364063 0.392188 0.034375 0.040625 |
0 0.438281 0.591406 0.195312 0.292187 |
0 0.366406 0.387500 0.029687 0.050000 |
0 0.439844 0.589063 0.195312 0.296875 |
0 0.371875 0.385937 0.021875 0.053125 |
0 0.439844 0.589844 0.195312 0.295312 |
0 0.365625 0.393750 0.034375 0.037500 |
0 0.440625 0.585938 0.190625 0.303125 |
0 0.365625 0.389844 0.031250 0.048438 |
0 0.437500 0.600000 0.196875 0.275000 |
0 0.366406 0.385156 0.032813 0.057813 |
0 0.537500 0.504687 0.406250 0.371875 |
0 0.309375 0.278906 0.034375 0.051562 |
0 0.517969 0.527344 0.104688 0.304688 |
0 0.327344 0.275000 0.026562 0.050000 |
0 0.514062 0.500000 0.125000 0.184375 |
0 0.252344 0.136719 0.026562 0.064062 |
0 0.513281 0.507812 0.114062 0.184375 |
0 0.266406 0.138281 0.026562 0.064062 |
0 0.511719 0.511719 0.110937 0.185938 |
0 0.264844 0.139844 0.026562 0.064062 |
0 0.510938 0.510938 0.109375 0.187500 |
0 0.266406 0.139844 0.026562 0.064062 |
0 0.522656 0.521094 0.126562 0.182812 |
0 0.278125 0.156250 0.025000 0.062500 |
0 0.520312 0.521875 0.128125 0.184375 |
0 0.272656 0.157812 0.026562 0.062500 |
0 0.519531 0.514844 0.126562 0.201563 |
0 0.273438 0.154688 0.025000 0.062500 |
0 0.518750 0.521875 0.125000 0.196875 |
0 0.275000 0.161719 0.025000 0.060937 |
0 0.517969 0.511719 0.117188 0.192188 |
0 0.282031 0.143750 0.026562 0.062500 |
0 0.515625 0.514062 0.112500 0.190625 |
0 0.281250 0.145313 0.025000 0.062500 |
0 0.532813 0.518750 0.140625 0.184375 |
0 0.289062 0.149219 0.025000 0.060937 |
0 0.535156 0.521875 0.139063 0.178125 |
0 0.297656 0.148438 0.023438 0.062500 |
0 0.528125 0.518750 0.134375 0.193750 |
0 0.290625 0.153125 0.025000 0.062500 |
0 0.535937 0.525000 0.146875 0.184375 |
0 0.296875 0.153906 0.025000 0.060937 |
0 0.532031 0.509375 0.139063 0.200000 |
0 0.295312 0.145313 0.025000 0.062500 |
0 0.508594 0.519531 0.132812 0.373437 |
0 0.245312 0.292969 0.021875 0.051562 |
0 0.751563 0.339844 0.065625 0.082812 |
0 0.510156 0.347656 0.092188 0.095312 |
0 0.477344 0.520312 0.448437 0.415625 |
0 0.658594 0.353125 0.020313 0.071875 |
0 0.435937 0.369531 0.031250 0.076563 |
0 0.977344 0.369531 0.045312 0.073438 |
0 0.510156 0.526563 0.160938 0.375000 |
0 0.595313 0.349219 0.028125 0.035937 |
0 0.366406 0.376563 0.039062 0.037500 |
0 0.865625 0.366406 0.053125 0.029687 |
0 0.494531 0.508594 0.148438 0.385937 |
0 0.475781 0.215625 0.029687 0.037500 |
0 0.193750 0.260156 0.053125 0.051562 |
0 0.725000 0.221094 0.046875 0.032813 |
0 0.507812 0.446875 0.187500 0.550000 |
0 0.627344 0.110937 0.032813 0.056250 |
0 0.593750 0.105469 0.028125 0.051562 |
0 0.526563 0.416406 0.218750 0.567187 |
0 0.563281 0.118750 0.032813 0.090625 |
0 0.279687 0.317188 0.021875 0.040625 |
0 0.243750 0.312500 0.021875 0.040625 |
0 0.501563 0.539844 0.146875 0.460938 |
0 0.257031 0.324219 0.026562 0.064062 |
0 0.176563 0.400000 0.021875 0.043750 |
0 0.142969 0.398438 0.032813 0.037500 |
0 0.502344 0.518750 0.095312 0.278125 |
0 0.185938 0.405469 0.025000 0.060937 |
0 0.149219 0.439063 0.023438 0.040625 |
0 0.099219 0.435156 0.032813 0.042188 |
0 0.505469 0.519531 0.064062 0.210938 |
End of preview. Expand in Data Studio
PASS Site-Adaptive Synthetic Data — Samples
2,000 fully labeled synthetic images sampled from the site-adaptive synthetic training datasets generated by the PASS (Passive-sensing Site Synthesis) pipeline — 500 images for each combination of construction site (small / large) and 3D background source (image-based 3D reconstruction / LiDAR scan).
Every image was rendered in a modified BlendCon engine using the real site's 3D model as the background, with automatically generated annotations — no manual labeling.
These are illustrative samples, not a benchmark. The complete synthetic datasets (tens of thousands of images per site/source) can be regenerated end-to-end with the open-source pipeline: github.com/SinaDavari/PaSS.
Dataset summary
| Subset | Site | Background source | Images | Boxes | Image size |
|---|---|---|---|---|---|
small_site/recon |
small | image-based 3D reconstruction | 500 | 1,157 | 640×640 |
small_site/lidar |
small | LiDAR scan (evaluation reference) | 500 | 824 | 640×640 |
large_site/recon |
large | image-based 3D reconstruction | 500 | 920 | 640×640 |
large_site/lidar |
large | LiDAR scan (evaluation reference) | 500 | 827 | 640×640 |
Each subset contains three aligned folders:
<site>/<source>/
├── images/ *.jpg # rendered synthetic images (Cycles, 640×640)
├── labels/ *.txt # YOLO format: 0 cx cy w h (normalized), class 0 = person
└── masks/ *.jpg # per-image semantic segmentation masks of the workers
Images and labels share the same filename stem; the corresponding mask
replaces the trailing test<NNNN> with Image<NNNN> (BlendCon convention).
Companion repositories
- Pipeline code: github.com/SinaDavari/PaSS
- Trained models (29 YOLOv10-s checkpoints):
SinaDavari/pass-worker-detection-yolov10s - Real annotated test sets:
SinaDavari/pass-worker-detection-test-sets
Intended use and limitations
- Intended use: inspecting the quality/characteristics of PASS-generated site-adaptive synthetic data; small-scale experimentation.
- Not a benchmark: subsets are random samples of much larger generated datasets; do not compare detectors on these.
- Site-specific by design: the data intentionally adapts to the two captured sites and is not meant to generalize.
License
Released under CC BY-NC 4.0 (attribution, non-commercial), consistent with the companion test sets.
Citation
Companion to the paper "PASS: Passive-sensing Site Synthesis" (submitted to Automation in Construction). Full citation to be added upon acceptance.
- Downloads last month
- 6