原书出处
Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)
A result of this analysis is we may explicit decompose contrastive learning into one loss for alignment and another for uniformity, with the following forms: (30.13) (30.14) (30.15) where ppos is the distribution of positive pairs and α, t, and λ are hyperparameters of the losses. 30.10.2 Experiment: Designing Embeddings with Contrastive Learning Using the alignment and uniformity objective defined previously, we will now illustrate
原书定位详情
来源:a19c7933-4eda-4176-aebd-30fe7e1ad292
正文位置:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c030b.xhtml
EPUB SHA-256:60a7eebc43f5d9a675d776fed3736fbdaa791e55877023f2f8cba016a9b384f8