논문 리뷰10 [논문 요약] ★초간단 5줄 요약★ ArcFace: Additive Angular Margin Loss for Deep Face Recognition / 얼굴인식 논문 ★초간단 5줄 요약★ ArcFace: Additive Angular Margin Loss for Deep Face Recognition ArcFace 논문의 목적얼굴 인식 성능 높이기 5줄 요약1. Additive Angular Margin Loss Term 추가2. 네트워크를 거쳐 도출된 feature embedding(vector)은 정규화를 통해 같은 hyperspher상에 위치.3. Inter-class 끼리는 feature vector의 geodesic distance 거리가 멀게, Intra-class 끼리는 sub-class로 더 구분하여 sub-class끼리의 거리가 멀게 학습.4. Sub-class 중 non-dominant cluster는 noise data.. 2024. 4. 24. Augmentation Matters: A Simple-yet-Effective Approach to Semi-supervisedSemantic Segmentation 논문 리뷰 Augmentation Matters: A Simple-yet-Effective Approach to Semi-supervised Semantic Segmentation https://arxiv.org/pdf/2212.04976.pdf - Github https://github.com/zhenzhao/AugSeg GitHub - ZhenZHAO/AugSeg: [CVPR'23] Augmentation Matters: A Simple-yet-Effective Approach to Semi-supervised Semantic Segmentati [CVPR'23] Augmentation Matters: A Simple-yet-Effective Approach to Semi-supervised Semantic S.. 2023. 7. 11. HET: Scaling out Huge Embedding Model Training via Cache-enabled Distributed Framework 논문 리뷰 빅데이터 분산 학습 프레임워크에 대한 논문이다. 빅데이터 과제의 일환으로 작성됨. 아카이브 링크: https://arxiv.org/abs/2112.07221 HET: Scaling out Huge Embedding Model Training via Cache-enabled Distributed Framework Embedding models have been an effective learning paradigm for high-dimensional data. However, one open issue of embedding models is that their representations (latent factors) often result in large parameter space. We obser.. 2023. 5. 17. [논문 리뷰] Accurate Facial Image Parsing at Real-Time Speed https://paperswithcode.com/paper/accurate-facial-image-parsing-at-real-time Papers with Code - Accurate facial image parsing at real-time speed#6 best model for Face Parsing on CelebAMask-HQ (Mean F1 metric)paperswithcode.com Accurate Facial Image Parsing at Real-Time Speed 논문 리뷰 Abstractpromising accuracy와 eal-time inference speed로 동작 가능한 face parsing 네트워크 제안일반 이미지 파싱과 face 파싱의 차이점을 분석하고 전통적인 F.. 2023. 4. 3. 이전 1 2 다음
