Visual Analysis and Perception Lab, Aalborg University

The Visual Analysis and Perception (VAP) Lab at Aalborg University has a strong history of collecting and curating datasets for computer vision in challenging, real-world scenarios.

We are slowly working on adding VAP Lab's datasets to this space to make them more accessible to the community. Below is an overview of dataset papers published by VAP Lab members in the past 5 years:

Underwater vision
Underwater Uncertainty: A Multi-Annotator Image Dataset for Benthic Habitat Classification (2024, ECCV Workshops)
BrackishMOT: The Brackish Multi-Object Tracking Dataset (2023, SCIA)
3D-ZeF: A 3D Zebrafish Tracking Benchmark Dataset (2020, CVPR)
Detection of Marine Animals in a New Underwater Dataset with Varying Visibility (2019, CVPR Workshops)

Thermal vision
Seasons in Drift: A Long-Term Thermal Imaging Dataset for Studying Concept Drift (2021, NeurIPS)
The Effect of a Diverse Dataset for Transfer Learning in Thermal Person Detection. (2020, Sensors)

Point cloud understanding
OpenTrench3D: A Photogrammetric 3D Point Cloud Dataset for Semantic Segmentation of Underground Utilities (2024, CVPR Workshops)
AssemblyNet: A Point Cloud Dataset and Benchmark for Predicting Part Directions in an Exploded Layout (2024, WACV)

Image enhancement
PDA-RWSR: Pixel-Wise Degradation Adaptive Real-World Super-Resolution (2024, WACV)
RELLISUR: A Real Low-Light Image Super-Resolution Dataset (2021, NeurIPS)

Traffic surveillance
Rain Removal in Traffic Surveillance: Does it Matter? (2019, IEEE Transactions on Intelligent Transportation Systems)
Is it Raining Outside? Detection of Rainfall using General-Purpose Surveillance Cameras. (2019, CVPR Workshops)

Fine-grained classification
Raw Instinct: Trust Your Classifiers and Skip the Conversion (2023, PRAI)
Sewer-ML: A Multi-Label Sewer Defect Classification Dataset and Benchmark (2021, CVPR)