{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreigz4ujsekc2zqdphszlojn323ghneudslv6innkdw4ivhnuve7ybq",
"uri": "at://did:plc:pgryn3ephfd2xgft23qokfzt/app.bsky.feed.post/3mgtyimlk37i2"
},
"path": "/t/feedback-request-synthetic-dataset-for-pedestrian-detection-in-extreme-night-fog/174215#post_1",
"publishedAt": "2026-03-12T08:17:54.000Z",
"site": "https://discuss.huggingface.co",
"tags": [
"JTSGRIT/NightFogPedestrianDataset_KiaRayPOV · Datasets at Hugging Face"
],
"textContent": "Hi everyone,\n\nI’ve recently published a synthetic dataset focused on a specific autonomous driving edge case: **Pedestrians in heavy night fog.** As we know, real-world data for extreme weather is often hard to collect and label. To address this, I used **Unity Perception** to simulate a night-time environment from the perspective of a vehicle (specifically a Kia Ray, with a camera height of 1.4m).\n\n**Key Features of this Dataset:**\n\n * **Environment:** High-density fog with low-light night conditions.\n\n * **Quantity:** 250 pairs of images and YOLOv8 formatted labels (Sample version).\n\n * **Perspective:** Realistic dashcam POV (Point of View).\n\n * **Variety:** Randomized pedestrian placements and fog densities.\n\n\n\n\n**I would highly appreciate your feedback on:**\n\n 1. **Visual Fidelity:** Does the synthetic fog and lighting look realistic enough to bridge the “Sim-to-Real” gap?\n\n 2. **Label Accuracy:** Are the bounding box annotations precise enough for your training pipelines?\n\n 3. **Data Value:** How useful do you find these types of synthetic edge cases for improving model robustness?\n\n\n\n\n**Dataset Link:** JTSGRIT/NightFogPedestrianDataset_KiaRayPOV · Datasets at Hugging Face\n\nI’m planning to expand this dataset to include more weather conditions (snow, heavy rain) and more classes (cyclists, motorcycles). Looking forward to your professional insights!\n\nBest regards,",
"title": "[Feedback Request] Synthetic Dataset for Pedestrian Detection in Extreme Night Fog"
}