Pedestrian perspective dataset improves urban obstacle detection accuracy
ROD-Dataset: A Pedestrian-Perspective Benchmark of Urban Obstacle Classes for YOLO-Based Detection
Computer Vision and Pattern RecognitionComputers and Society
Summary
People walking in cities can be in danger when they encounter obstacles they do not see in time, especially those with vision issues or distracted walkers. The authors provide a large set of labeled images taken from a pedestrian's view to train AI systems to spot common street obstacles like stairs, benches, and manholes. They combined data from many sources and cities to make the dataset more useful across locations. They also tested several lightweight AI models and found that some versions of YOLO detectors can identify most obstacles reliably. This helps build warning systems to alert pedestrians earlier about hazards.
What this means in practice
- •For assistive device makers: Train pedestrian-level obstacle detection systems that improve timely hazard alerts for visually impaired and distracted pedestrians.$Commercial implications: Enables development of wearable or phone-based assistive devices that warn users about urban obstacles to enhance safety.
- •For mobile app developers: Build pedestrian safety apps that detect common sidewalk hazards with accurate, lightweight models tested on a diverse dataset.
Authors
Abtin Zandi, Ariyan Azami, Bardia Sabagh Kermani, Parsa Abbasian, Roza Ganjipour, Sarvin Nami, Hamed Farbeh
Abstract
Two groups of pedestrians face avoidable risk on the city sidewalk: people living with visual impairment, who reach an obstacle before anything warns them, and people walking with divided attention, who register a hazard late or not at all. A vision system can warn them only about obstacles it has learned to recognize, so its training data limits the protection it offers. No suitable training set exists: public benchmarks are recorded from vehicles, aerial platforms, or fixed rigs, general-purpose detectors omit the street furniture that causes trips, and pedestrian-focused collections are confined to single regions with incompatible class schemes. This paper introduces the Real-time Obstacle Detection (ROD) dataset: more than 24,000 images and 40,000 annotated obstacles across 25 classes, from vehicles and people to the stairs, manholes, traffic cones, and benches that general-purpose detectors overlook. Because street furniture differs between cities, ROD unifies twenty-nine public collections onto one taxonomy and adds eye-level imagery from two contrasting cities on two continents, Tehran and Toronto. Every label is human-verified, and each capture campaign is admitted only after a compact detector shows it can learn the class. Six lightweight YOLO detectors (nano and tiny variants) benchmarked on ROD all exceed 0.86 recall; YOLO12n attains the highest recall (0.889), YOLO26n the highest precision (0.925), and YOLOv9t the best mAP@0.50 (0.882) with the fewest parameters. Because the recall and mAP rankings disagree, selecting a model by mAP does not minimize missed obstacles, so per-class recall is reported for all 25 classes.