論文閱讀 (ECCV 2016),cited:1530
https://github.com/richzhang/colorization
Contribution
設計loss function來處理著色問題中的multimodal (單一物體可以有多種上色可能),使上色有多樣性及合理性,而結果是否為ground truth則不是重點。
For this paper, our goal is not necessarily to recover the actual ground truth color, but rather to produce a plausible colorization that could potentially fool a human observer.
上色任務轉換為self-supervised feature learning,並在一些benchamrks中達到頂尖水平。
Jian-Chang changed 4 years agoView mode Like Bookmark