MovingMNIST(root)ĭatasetFolder(root, loader)Ī generic data loader where the images are arranged in this way by default. UCF101(root, annotation_path, frames_per_clip) HMDB51(root, annotation_path, frames_per_clip) Multi-view Stereo Correspondence Dataset.ĬocoCaptions(root, annFile)

Publicly available scenes from the Middlebury dataset 2014 version. SceneFlowStereo(root)ĭataset interface for Scene Flow datasets. Synthetic dataset used in training the CREStereo architecture.įallingThingsStereo(root) KITTI dataset from the 2015 stereo evaluation benchmark. KITTI dataset from the 2012 stereo evaluation benchmark. Sintel(root)Ĭarla simulator data linked in the CREStereo github repo. SUN397(root)ĬocoDetection(root, annFile)Ĭityscapes(root) German Traffic Sign Recognition Benchmark (GTSRB) Dataset.

Large-scale CelebFaces Attributes (CelebA) Dataset Dataset.Ĭountry211(root)ĭTD(root)ĮuroSAT(root)įakeData()Ī fake dataset that returns randomly generated images and returns them as PIL imagesįashionMNIST(root)įlickr8k(root, ann_file)įlickr30k(root, ann_file)įlowers102(root) You can also create your own datasets using the provided base classes. Transform and target_transform to transform the input and target respectively. nThreads )Īll the datasets have almost similar API. DataLoader ( imagenet_data, batch_size = 4, shuffle = True, num_workers = args. ImageNet ( 'path/to/imagenet_root/' ) data_loader = torch.

8 section classification folder