URL details: yash0307.github.io/RecallatK_surrogate
URL title:
Recall@k Surrogate Loss with Large Batches and Similarity Mixup
URL paragraphs:
Yash Patel Giorgos Tolias Jiří Matas Direct optimization, by gradient descent, of an evaluation metric, is not possible when it is non-differentiable, which is the case for recall in retrieval. In this work, a differentiable surrogate loss
URL keywords:
Learning Surrogates, Recall@k, Recall@k Surrogate Loss, Mixup, Similarity Mixup, Non-Differentiable Losses
URL last crawled:
2023-01-07
URL speed:
147.000 MB/s,
downloaded in 0.001 seconds
We found no external links pointing to this url.