Reproduce “A2D2: Audi Autonomous Driving Dataset” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jakob Geyer, Yohannes Kassahun, Mentar Mahmudi, Xavier Ricou, Rupesh Durgesh, Andrew S. Chung, Lorenz Hauswald, Viet Hoang Pham and 11 more
A2D2 provides aligned surround-view camera, LiDAR, and vehicle-state data, with annotated perception examples and longer unannotated sequences. The authors describe its emphasis on semantic segmentation, vehicle-bus signals, and sequence data alongside other multimodal datasets. A ResNet-101 with PSP-Net reaches 71.01% mean IoU on the test set; anonymized images reach 70.94%, in this segmentation setup only.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.