Loading...
Loading...
Self-driving perception, LiDAR/camera fusion, trajectory prediction, drone perception, robot manipulation
2,023 datasets
UrbanLoco provides a multi-sensor suite for mapping and localization in urban canyons, developed by Weisong Wen and updated in 2025. The data includes LiDAR and camera streams designed for autonomous vehicle navigation in dense city environments where GNSS signals are frequently obstructed. While the exact record count is not specified, the dataset is tailored for SLAM and mapping research.
7,465 question-answer pairs in Egyptian Arabic cover topics in Islamic studies. The dataset was created by Omar-youssef and last updated in October 2025. It serves as a resource for developing Arabic natural language processing models focused on religious knowledge.
L2D is a large-scale autonomous driving dataset created by yaak-ai and hosted on Hugging Face. It contains over 90 terabytes of multimodal data, representing more than 5,000 hours of driving footage collected from 30 cities in Germany. The dataset was last updated on September 29, 2025.
Massive-STEPS-Jakarta is a large-scale dataset of semantic trajectories derived from Foursquare Open Source Places and the Semantic Trails Dataset. It contains check-in data from Jakarta, Indonesia, and is intended for research in trajectory prediction and urban modeling. The dataset was created by the CRUISE Research Group and was last updated in September 2025.
A dataset uploaded to HuggingFace by user humair025 on 2025-11-09. The title suggests it contains Urdu language content related to Islam. The dataset's specific content, size, and structure are not detailed in the provided metadata.
A catalog of Spanish-authored artistic and cultural works linked to the Holocaust from 1933 to 2024. The dataset, created by María Jesús Ferrnández-Gil, allows identification by typology, creation date, language, promoting institution, and author. It was last updated on October 14, 2025.
Presenting a single log from the NuPlan autonomous driving dataset, converted into the VADv2 training and evaluation format. It contains sensor data including camera images and fused LiDAR point clouds, alongside a map bundle and a SQLite database for the mini split.
Griffin is a pioneering publicly available dataset for aerial-ground cooperative 3D perception. It features over 200 dynamic scenes, totaling more than 30,000 frames and 270,000 images, built using CARLA-AirSim co-simulation. The dataset was created by author wjh-svm and last updated on Hugging Face in September 2025.
Game4Loc is a UAV geo-localization benchmark created by Yux1angJi from synthetic game data, featured as an AAAI 2025 Oral presentation. The dataset provides visual imagery and geographic coordinates for visual place recognition and localization tasks, updated as of October 2025.
AI4MARS provides ~326,000 semantic segmentation labels for 35,000 images from NASA's Curiosity, Opportunity, and Spirit rovers. The dataset was created for training terrain classification models to support autonomous driving on Mars, with labels collected via crowdsourcing and validated by NASA mission scientists. Each crowdsourced image label was produced by 10 annotators to ensure quality.
U2UData-2 is a large-scale simulation dataset for swarm UAV autonomous flight, specifically designed for Long-Horizon tasks. It was created by author fengtt42 and last updated on the Hugging Face platform in September 2025. The dataset addresses limitations in existing methods by modeling complex dependencies and dynamic goal shifts required for real-world deployment.
U2UData-2 is a large-scale dataset for swarm Unmanned Aerial Vehicle autonomous flight focused on Long-Horizon tasks. It was created by author fengtt42 to address limitations in existing methods that fail in real-world deployment for complex, non-sequential missions. The dataset was last updated in September 2025.
UAVScenes is a large-scale dataset introduced for ICCV 2025 to benchmark tasks across 2D and 3D modalities. It is built upon the MARS-LVIG dataset and enhanced with manually labeled semantic annotations for images and LiDAR point clouds. The dataset was uploaded to Hugging Face by user sijieaaa on August 6, 2025.
Waymo Open Dataset V 1.4.3 is a collection of sensor data for autonomous driving research, released by Waymo. It contains data for tasks like motion prediction, object detection, and computer vision. The dataset was uploaded to HuggingFace by AnnaZhang in September 2025.
Waymo Open Dataset Motion V 1.3.0 provides data for predicting the movement of road users. It is released by Waymo and was updated on HuggingFace in September 2025. The dataset is designed for tasks like motion forecasting and trajectory modeling.
A subsample of the Waymo Open Dataset, governed by its specific license agreement. The dataset contains image, video, and sensor data for traffic surveillance and computer vision tasks, contributed by author mickeykang and last updated in August 2025.
Published at ACM Multimedia 2024 in Melbourne, Australia, this dataset is designed for haze-aware single image dehazing. It was created by researchers from Sungkyunkwan University, ANU, CSIRO, and NIT Rourkela. A satellite version of the dataset was uploaded in July 2025.
NVIDIA provides a multimodal dataset for autonomous vehicle research, last updated on June 15, 2025. The dataset includes synthetic data, HD maps, and LiDAR point clouds. A download script is available for users with sufficient storage space.
RFUAV contains 1.3 TB of raw radio-frequency (RF) signal data collected from 37 distinct unmanned aerial vehicle (UAV) models for detection and identification tasks. Developed by kitofrank and released in 2025, this benchmark provides high-volume signal data to address the limitations of smaller, less diverse drone signal repositories.