Processed Dataset Files

Changes to the previous version

• Add the new split univ13 in ETH-UCY, which includes training sets {eth, hotel, unive, zara1, zara2, zara3} and test sets {univ, univ3};
• Fix the type typo in ETH-UCY: Pedestrain -> Pedestrian;
• Fix the wrong split rate in nuScenes_ov, which now contains 550 training sets.

Usages

Please unzip and put them into the corresponding dataset folders (dataset_configs and dataset_processed) to the root path of the training codes, like

____ROOT
    |____dataset_configs
        |____ETH-UCY
            |____subsets
            |____eth.plist
            |____hotel.plist
            |____ ...
    |____dataset_processed
        |____ETH-UCY
            |____eth
            |____hotel
            |____ ...
    |____main.py
    |____ ...