Sign in to view source links and access this dataset
Description
OctoSense is a time-synchronized, calibrated multi-sensor driving dataset. It comprises 371 sequences totaling 59.0 hours and 2,474 km of driving across urban, suburban, and rural areas on Long Island and in Philadelphia, covering sunrise, daytime, sunset, and nighttime conditions. The dataset, created by anthonytec2, includes synchronized stereo RGB, stereo event cameras, infrared, Ouster LiDAR, two IMUs, GPS, CAN data, and ground truth annotations.
Use Cases
Develop sensor fusion models based on synchronized stereo RGB, event cameras, infrared, and LiDAR data.
Train and validate ego-motion estimation algorithms based on ground truth optical flow and IMU data.
Benchmark depth estimation models based on the provided ground truth depth data.
Study autonomous driving perception across diverse lighting conditions (sunrise, daytime, sunset, nighttime) and road types (highway, residential, city).
Research multi-modal localization and mapping using synchronized GPS, CAN, and LiDAR data.
Strengths
Large scale with 371 sequences, 59.0 hours of driving, and 2,474 km covered.
Multi-sensor suite includes synchronized stereo RGB, stereo event cameras, infrared, Ouster LiDAR, two IMUs, GPS, and CAN data.
Diverse environmental coverage across urban, suburban, and rural areas in two distinct geographic regions (Long Island and Philadelphia).
Captures varied lighting conditions including sunrise, daytime, sunset, and nighttime.
Limitations
Description metadata is limited; actual data quality requires manual inspection after download.
Column-level documentation is absent; field semantics must be inferred after download.
Data may reflect geographic and temporal bias inherent to the specific collection locations and times.
Provenance
Source
huggingface user anthonytec2
Collection Method
Likely collected using a sensor-equipped vehicle during driving sequences.
Time Range
Sequences span sunrise, daytime, sunset, and nighttime conditions.
Freshness
Last updated 2026-06-12 11:03:54; freshness should be verified.
Geography
Long Island and Philadelphia, USA, covering urban, suburban, and rural driving environments.
License is unknown; users must verify terms of use before downloading.