https://arxiv.org/abs/2206.07669
A Unified Sequence Interface for Vision Tasks (Ting Chen, Saurabh Saxena, Lala Li, Tsung-Yi Lin, David J. Fleet, Geoffrey Hinton)
예상하던 그것이 나왔네요. multitask pix2seq. instance segmentation도 좌표 예측으로 통합이 가능하군요.
#multitask #object_detection #instance_segmentation #keypoint