MOTOR: Multi-View Two-Wheeler Rider Behavior in Dense Indian Traffic

India is the source for this dataset of two-wheeler rider behavior in dense, unstructured traffic. The full dataset comprises 1,629 annotated sequences (~25 hours) from 16 riders, collected across diverse traffic scenarios. It was created by Voxel51 and is a multi-view, multimodal dataset.

Use Cases

• Train behavior prediction models based on multi-view video sequences.
• Develop anomaly detection systems for unstructured traffic conditions.
• Benchmark multimodal (likely video + sensor) fusion algorithms for rider analysis.
• Study traffic patterns and safety in dense urban environments.

Strengths

• 1,629 annotated sequences provide a substantial sample size.
• Data collection involved 16 distinct riders, suggesting diversity in behavior.
• The dataset covers ~25 hours of observation across diverse traffic scenarios.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The repository contains only a subset of the full MOTOR dataset.

Provenance

Source

Voxel51

Collection Method

Collected from 16 riders across diverse traffic scenarios.

Freshness

Last updated 2026-05-26 20:33:38; freshness should be verified.

Geography

India