Temporal Transductive Inference for Few-Shot Video Object Segmentation


Online demo tool soon ...



Please specify the base classes and upload the support set image-label pairs.

Training Classes

Select from four folds, each fold contain certain set of classes already trained on (base classes).



Base Classes List:



Support Set

Upload Images for the novel class or Use sample.

Upload Corresponding Masks if you have uploaded the images.

If using a sample no masks are required.

Please upload the test video frames Or use a sample then press submit.