Index
- Index
- Metric Learning / Distance Learning
- 損失 / Loss
- NCA Loss / 2014
- Contrastive Loss / 2005
- Triplet Loss / Triplet Margin Loss / 2005
- Ranking Loss / 2013
- Lifted Structure Loss / 2015
- Multi-Class N-Pair Loss / 2016
- Large Margin Softmax Loss / 2016
- Angular Loss / 2017
- Generalized Lifted Structure Loss / 2017
- Instance Loss / 2017
- Margin Loss / 2017
- Proxy NCA Loss / Proxy Ranking Loss / 2017
- Normalized Softmax Loss / 2018
- Ranked List Loss / 2019
- Fast AP Loss / 2019
- Multi Similarity Loss / 2019
- Signal To Noise Ration Contrastive Loss / 2019
- Soft Triplet Loss / 2019
- Tuplet Loss / Tuplet Margin Loss / 2019
- Intra Pair Variance Loss / 2019
- Circle Loss / 2020
- Proxy Anchor Loss / 2020
- Sup Con Loss / 2020
- Centroid Triplet Loss / 2021
- VIC Reg Loss / 2021
- Face Recognition
Metric Learning / Distance Learning
- Metric Learning / Distance Learning #まとめ編 #00
ここでは、特に損失について、記す.
損失 / Loss
NCA Loss / 2014
マハラノビス距離を導入.
マハラノビス距離
Neighbourhood Components Analysis
- [2014]
- papers.nips.cc
Contrastive Loss / 2005
「異なる」種類のデータ間の距離を大きく (遠く)、「同じ」種類のデータ間の距離を小さく (近く)
なるように写像することを目的とした Loss.
- Contrastive Loss
Triplet Loss / Triplet Margin Loss / 2005
Triplet Architecture / Triplet Network で利用される.
- Triplet Architecture / Triplet Network
Ranking Loss / 2013
“Devise: A deep visual-semantic embedding model,” in NIPS, 2013.
“Deep fragment embeddings for bidirectional image sentence mapping,” in NIPS, 2014.
“Multimodal convolutional neural networks for matching image and sentence,” in ICCV, 2015.
“Learning deep structure-preserving image-text embeddings,” in CVPR, 2016.
“Dual attention networks for multimodal reasoning and matching,” in CVPR, 2017.
“Learning deep representations of fine-grained visual descriptions,” in CVPR, 2016.
Lifted Structure Loss / 2015
- Lifted Structure Loss
Multi-Class N-Pair Loss / 2016
Tuplet Loss と Multi-Class N-Pair Loss.
- Multi-Class N-Pair Loss
NTXent Loss / 2018
N-Pair Loss の一般化.
Info NCE とも呼ばれる.
Representation Learning with Contrastive Predictive Coding
- [2018]
- arxiv.org
Momentum Contrast for Unsupervised Visual Representation Learning
- [2019]
- arxiv.org
A Simple Framework for Contrastive Learning of Visual Representations
- [2020]
- arxiv.org
Large Margin Softmax Loss / 2016
- Large-Margin Softmax Loss for Convolutional Neural Networks
- [2016]
- arxiv.org
Angular Loss / 2017
- Deep Metric Learning with Angular Loss
- [2017]
- arxiv.org
Generalized Lifted Structure Loss / 2017
Person Re-Identification の研究.
-
- Person Re-Identification
- Person Re-Identification
In Defense of the Triplet Loss for Person Re-Identification
- [2017]
- arxiv.org
Instance Loss / 2017
Image と Text の Embedding を用いる研究.
- Dual-Path Convolutional Image-Text Embeddings with Instance Loss
- [2017]
- 4 PROPOSED INSTANCE LOSS
- 4.2 Instance Loss
- arxiv.org
Margin Loss / 2017
データのサンプリングについての研究.
- Sampling Matters in Deep Embedding Learning
- [2017]
- arxiv.org
Proxy NCA Loss / Proxy Ranking Loss / 2017
- No Fuss Distance Metric Learning using Proxies
- [2017]
- 3 Metric Learning using Proxies
- 3.2 Proxy Ranking Loss
- arxiv.org
Normalized Softmax Loss / 2018
- Classification is a Strong Baseline for Deep Metric Learning
- [2018]
- arxiv.org
Ranked List Loss / 2019
- Ranked List Loss for Deep Metric Learning
- [2019]
- v8
- arxiv.org
Fast AP Loss / 2019
- Deep Metric Learning to Rank
Multi Similarity Loss / 2019
- Multi-Similarity Loss with General Pair Weighting for Deep Metric Learning
- [2019]
- v3
- arxiv.org
Signal To Noise Ration Contrastive Loss / 2019
- SNR Contrastive Loss
Soft Triplet Loss / 2019
従来の Softmax Loss より改善した.
- Soft Triplet Loss
Tuplet Loss / Tuplet Margin Loss / 2019
Batch からのサンプリングについての研究.
- Deep Metric Learning with Tuplet Margin Loss
- [2919]
- 3 Method
- 3.3 Tuplet Margin Loss
- https://openaccess.thecvf.com/content_ICCV_2019/papers/Yu_Deep_Metric_Learning_With_Tuplet_Margin_Loss_ICCV_2019_paper.pdf
Intra Pair Variance Loss / 2019
Batch からのサンプリングについての研究.
- Deep Metric Learning with Tuplet Margin Loss
- [2019]
- 3 Method
- 3.4 Intra-pair Variation
- https://openaccess.thecvf.com/content_ICCV_2019/papers/Yu_Deep_Metric_Learning_With_Tuplet_Margin_Loss_ICCV_2019_paper.pdf
Circle Loss / 2020
- Circle Loss: A Unified Perspective of Pair Similarity Optimization
- [2020]
- arxiv.org
Proxy Anchor Loss / 2020
- Proxy Anchor Loss for Deep Metric Learning
- [2020]
- arxiv.org
Sup Con Loss / 2020
- Supervised Contrastive Learning
- [2020]
- v5
- arxiv.org
Centroid Triplet Loss / 2021
- Centroid Triple Loss
VIC Reg Loss / 2021
- VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning
- [2021]
- v3
- arxiv.org
Face Recognition
Face Recognition で利用される損失.
Sphere Face Loss / 2017
- SphereFace: Deep Hypersphere Embedding for Face Recognition
- [2017]
- arxiv.org
Arc Face Loss / 2018
- ArcFace
Cos Face Loss / 2018
Sub Center Arc Face Loss / 2020
- Sub-center ArcFace
