Projection Head
رأس الإسقاط
شبكة صغيرة (عادةً MLP) تُضاف فوق الهيكل الأساسي لتحويل السمات إلى فضاء تضمين منخفض الأبعاد تُحسب فيه الخسارة التبايُنية.
A small network (typically MLP) added on top of the backbone to project features into a lower-dimensional embedding space where the contrastive loss is computed.
Also translated asطبقة الإسقاط، رأس التحويل
First appears in this corpus in: Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning (2020)
Appears in these papers
- Barlow Twins: Self-Supervised Learning via Redundancy Reduction2021in the sky ✦
- Barlow Twins: Self-Supervised Learning via Redundancy Reduction2021in the sky ✦
- Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning2020in the sky ✦
- Emerging Properties in Self-Supervised Vision Transformers2021in the sky ✦
- Emerging Properties in Self-Supervised Vision Transformers2021in the sky ✦
- ImageBind: One Embedding Space to Bind Them All2023in the sky ✦
- ImageBind: One Embedding Space to Bind Them All2023in the sky ✦
- A Simple Framework for Contrastive Learning of Visual Representations2020in the sky ✦
- A Simple Framework for Contrastive Learning of Visual Representations2020in the sky ✦
- Exploring Simple Siamese Representation Learning2021in the sky ✦
- Exploring Simple Siamese Representation Learning2021in the sky ✦
- SwAV: Unsupervised Learning of Visual Features by Contrasting Cluster Assignments2020in the sky ✦
- VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning2021in the sky ✦