[논문리뷰] Towards a Definition of Disentangled Representations

[논문리뷰] Towards a Definition of Disentangled Representations

항목내용저자Irina Higgins, David Amos, David Pfau, Sebastien Racaniere, Loic Matthey, Danilo Rezende, Alexander Lerchner연도2018 (December)제목Towards a Definition of Disentangled Representations학회.소속Google DeepMind (DeepMind) Disentangled Representation Learning(분리 표현 학습)Robustness, Generalisability 를 위해 dataset, model architecture 등을 수정하는 다양한 시도들이 있었음아키텍처를 고정하는 대신, 데이터의 구조를 잘 반영하는 표현(Representation)을 학..

  • format_list_bulleted Paper review/Disentanglement
  • · 2025. 12. 15.
[논문리뷰] Towards Better Disentanglement in Non-Autoregressive Zero-Shot Expressive Voice Conversion

[논문리뷰] Towards Better Disentanglement in Non-Autoregressive Zero-Shot Expressive Voice Conversion

1. Bibliographic Info항목내용저자Seymanur Akti, Tuan Nam Nguyen, Alexander Waibel연도2025제목Towards Better Disentanglement in Non-Autoregressive Zero-Shot Expressive Voice Conversion학회Interspeech 2025 (Rotterdam, Netherlands)소속Karlsruhe Institute of Technology, Carnegie Mellon University2. Problem StatementEVC : expressive voice conversion기존의 목소리 변환(Voice Conversion, VC) 기술에서 한 단계 더 나아가, '감정', '억양', '운율(..

  • format_list_bulleted Paper review/Disentanglement
  • · 2025. 12. 15.

Spectral Theorem : ML

일반적인 Eigen Vecter들은 서로 Orthogonal 할 필요가 없음비스듬히 미는 변환(Shear)에서 변하지 않는 축(고유벡터)들이 90도가 아니라 30도나 45도로 좁게 모여 있을 수도 있다. Linearly Independent는 맞지만 Orthogonal일 필요는 없음. 하지만 Symmetric Matrix 을 만족하는 경우 항상 Orthogonal 하며 이를 Spectral Theorem 이라 부름 Spectral Theorem [Linear Algebra] Lecture 25 대칭 행렬(Symmetric Matrix)과 스펙트럼 정리(Spectral Theorem)이번 강의에서는 대칭 행렬(Symmetric Matrix)에 대해 이야기 하도록 하겠다. 지난 강의 에서 간략히 배우긴 했..

  • format_list_bulleted AI/ML basic
  • · 2025. 12. 11.

Eigenvalues, Eigenvectors : ML

Vector\begin{bmatrix}x \\y \\z\end{bmatrix} 원점(0,0)에서 출발하는 화살표공간상의 **한 지점(Point)**을 콕 찍는 것 Matrix\begin{bmatrix} a & \cdots & b \\ \vdots & \ddots & \vdots \\ c & \cdots & d \end{bmatrix}모눈종이(Grid) 전체를 움직이는 함수행렬을 곱한다는 것은 공간을 찌그러뜨리고, 늘리고, 회전시키는 행위 Determinant $$ \mathrm{det}(A) = ad - bc $$ 원래 넓이 1이었던 정사각형이 변형 후에 넓이가 얼마가 되었는지를 계산하는 것 Eigenvectors, Eigenvalues$$ Ax = \lambda x $$좌변: 행렬 A가 벡터 x를 ..

  • format_list_bulleted AI/ML basic
  • · 2025. 12. 11.
[논문리뷰] Disentanglement network: Disentangle the emotional features from acoustic features for speech emotion recognition

[논문리뷰] Disentanglement network: Disentangle the emotional features from acoustic features for speech emotion recognition

1. Bibliographic Info항목내용저자Zhichen Yuan, C. L. Philip Chen, Shuzhen Li, Tong Zhang연도2024제목Disentanglement Network: Disentangle the Emotional Features from Acoustic Features for Speech Emotion Recognition학회ICASSP 20242. Introduction and Background2.1 Application Areas음성 비서, 고객 서비스, 의료 로봇 등 인간-컴퓨터 상호작용(HCI) 시스템에서 사용자 감정에 맞춤화된 응답 제공에 활용 가능.2.2 Related Work연구접근법Peri et al. (2021)화자 인식(SR)을 보조 태스크로 사..

  • format_list_bulleted Paper review/Disentanglement
  • · 2025. 12. 11.
[논문리뷰] Simple Disentanglement of Style and Content in Visual Representations

[논문리뷰] Simple Disentanglement of Style and Content in Visual Representations

1. Bibliographic InfoAuthors: Lilian Ngweta, Subha Maity, Alex Gittens, Yuekai Sun, Mikhail YurochkinYear: 2023Title: Simple Disentanglement of Style and Content in Visual RepresentationsConference: ICML 2023 (International Conference on Machine Learning)2. Introduction and Background2.1 Application Areas:out-of-distribution (OOD) generalizationimage retrievalimage-to-image translationvisually-a..

  • format_list_bulleted Paper review/Disentanglement
  • · 2025. 12. 4.
[논문리뷰] Do Audio-Language Models Understand Linguistic Variations?

[논문리뷰] Do Audio-Language Models Understand Linguistic Variations?

1. Bibliographic Info항목내용저자Ramaneswaran Selvakumar, Sonal Kumar, Hemant Kumar Giri, Nishit Anand, Ashish Seth, Sreyan Ghosh, Dinesh Manocha연도2025제목Do Audio-Language Models Understand Linguistic Variations?출처Accepted to NAACL 2025소속University of Maryland, College Park; NVIDIA, Bangalore Do Audio-Language Models Understand Linguistic Variations?Ramaneswaran Selvakumar, Sonal Kumar, Hemant Kumar G..

  • format_list_bulleted Paper review/Audio Language Model
  • · 2025. 12. 4.
[논문리뷰] Contrastive Learning Inverts the Data Generating Process

[논문리뷰] Contrastive Learning Inverts the Data Generating Process

1. Bibliographic InfoAuthors: Roland S. Zimmermann, Yash Sharma, Steffen Schneider, Matthias Bethge, Wieland BrendelYear: 2021Title: Contrastive Learning Inverts the Data Generating ProcessConference: ICML 2021 (38th International Conference on Machine Learning) 2. Introduction and Background2.1 Application AreasSelf-supervised representation learning for visual and sequential dataImage classifi..

  • format_list_bulleted Paper review/Disentanglement
  • · 2025. 12. 3.
[논문리뷰] Towards LLM-Empowered Fine-Grained Speech Descriptors for Explainable Emotion Recognition

[논문리뷰] Towards LLM-Empowered Fine-Grained Speech Descriptors for Explainable Emotion Recognition

1. Bibliographic InfoAuthors: Youjun Chen, Xurong Xie, Haoning Xu, Mengzhe Geng, Guinan Li, Chengxi Deng, Huimeng Wang, Shujie Hu, Xunying LiuYear: 2025Title: Towards LLM-Empowered Fine-Grained Speech Descriptors for Explainable Emotion RecognitionVenue: Accepted by INTERSPEECH2025Institution: The Chinese University of Hong Kong, Institute of Software CAS, National Research Council Canada 2. Int..

  • format_list_bulleted Paper review/Audio Language Model
  • · 2025. 12. 1.
[논문리뷰] Towards Disentangled Speech Representations

[논문리뷰] Towards Disentangled Speech Representations

1. Bibliographic InfoAuthors: Cal Peyser, Ronny Huang, Andrew Rosenberg, Tara N. Sainath, Michael Picheny, Kyunghyun ChoAffiliations: Center for Data Science, New York University; Google Inc.Year: 2022Venue: Interspeech 2022 2. Introduction and Background2.1 Application AreasSemi-supervised ASR 시스템: Back-transcription 기반 시스템(Speech Chain, Sequential MixMatch)의 개선Unsupervised Pretraining: 음성 도메인에..

  • format_list_bulleted Paper review/Disentanglement
  • · 2025. 12. 1.
[논문리뷰] DECODEC: Rethinking audio codecs as universal desentangled representation learners

[논문리뷰] DECODEC: Rethinking audio codecs as universal desentangled representation learners

1. Bibliographic Info항목내용저자Xiaoxue Luo, Jinwei Huang, Runyan Yang, Yingying Gao, Junlan Feng, Chao Deng, Shilei Zhang연도2025제목DeCodec: Rethinking Audio Codecs as Universal Disentangled Representation Learners출처arXiv:2509.09201v1 (2025.09.11)2. Introduction and Background 음성을 3가지로 분리한 codec ( Disentanglement )semantic speechparalinguistic speechbackground sound 2.1 Application AreasSpeech Enhance..

  • format_list_bulleted Paper review/Codec
  • · 2025. 12. 1.

내가 보려고 만든 도커 명령어 모음집

1. Docker 기본 개념🎯 핵심 구성요소이미지: 실행 파일 + 라이브러리 + 설정 (읽기 전용)컨테이너: 이미지를 실행한 인스턴스 (읽기/쓰기)볼륨: 데이터 영속성을 위한 저장소네트워크: 컨테이너 간 통신2. Docker 이미지 관리📦 이미지 조회# 모든 이미지 목록docker imagesdocker image ls# 특정 이미지 검색docker images | grep pythondocker images python# 이미지 상세 정보docker inspect python:3.11# 이미지 히스토리 (레이어 확인)docker history python:3.11# 이미지 크기 확인docker images --format "table {{.Repository}}\t{{.Tag}}\t{{.Size}}..

  • format_list_bulleted CS/Development
  • · 2025. 11. 14.