Multi-Attention Network for Compressed Video Referring Object Segmentation

Benchmark Model Rank Results
referring-expression-segmentation-on-a2d-sentencesMANET#5AP: 0.471IoU overall: 0.726IoU mean: 0.632Precision@0.5: 0.734…
referring-expression-segmentation-on-refer-youtube-vos-2021-public-validationMANET#41J&F: 55.63J: 54.75F: 56.51