Minha Kim[Paper Review]MobileNetV4-Universal Models for the Mobile Ecosystem (Google research)I will introduce about On-device model search paper, MobileNetV4-Universal Models for the Mobile Ecosystem (Google research)Jul 16Jul 16
Minha Kim[Paper Review] Stitchable Neural Networks (CVPR’23-Highlight)This approach is suitable for various deployment scenarios, enabling faster training and improved performance.Jun 12Jun 12
Minha Kim[Paper Review] RAP-SAM: Towards Real-Time All-Purpose Segment Anything“RAP-SAM: Towards Real-Time All-Purpose Segment Anything.” from Google ResearchApr 24Apr 24
Minha Kim[Paper Review] Segment Anything (arXiv ’23, Meta Research)the paper GPT of segmentation from META reseracher.Feb 17Feb 17
Minha Kim[Paper Review] AAAI’23 — Learning Polysemantic Spoof Trace: A Multi-Modal Disentanglement Network…AbstractJan 23Jan 23
Minha KimWhat is Active learning?Active Learning is one of the methods to enhance machine learning models by effectively selecting and utilizing data. There are various…Jan 17Jan 17
Minha Kim[Analytics] Kernel Function & SpaceWhat Kernel Function? And How is this function used?Jan 10Jan 10
Minha KimLoss Functions for Semantic Segmentation최근 Segmentation쪽으로 업무를 진행하면서, Loss function을 이론적으로 정리를 해보고자 하는 마음이 생겼다. 대부분 Dice, CE Loss를 적용하고 있는데, 많이 알아둘 수록 combination Loss로 진행해보고…Jan 9Jan 9
Minha Kim[Paper-review] Meta-Learned Confidence for Few-shot Learning (CVPR2020)IntroductionJan 4Jan 4