changjinhan
EXPLORE THE ARCHIVE

Browse by tags.

Find something to read in a topic that interests you.

Agent · 7Agentic RL · 1AI Agent · 1AI Ops · 1Anti Spoofing · 2ASR · 2Audio Augmentation · 1Audio Compression · 2Audio Deepfake Detection · 2Audio Synthesis · 1Auto Harness · 1BERT · 1Byte representation · 1Code Switching · 1Codec · 2Continual Learning · 1Contrastive Learning · 1Cross-Lingual TTS · 1Custom Voice · 3Diffusion · 1DLM · 1Duration · 1Dynamic Time Warping · 1Finance · 2Financial Engineering · 1Flow · 2Flow Matching · 1Full Stack Agent · 1G2P · 8GAN · 5GNN · 1HACRL · 1Hugging Face · 1Lexicon · 1LLM · 5Long Sequence · 1LoRA · 1Loss Function · 1LSTM · 2Mamba3 · 1Meta Learning · 1MetaClaw · 1MoE · 1MOS · 3Multi Speaker · 1Natural Language Guidance · 1Neural Codec Language Model · 4Neural Vocoder · 1On-Device AI · 1Online Course · 1OpenSeeker · 1Positional Embedding · 1Privacy · 1Prompt TTS · 1Quantization · 1Research Automation · 1RL · 1Self Evolution · 1Seq2Seq · 1SER · 1Skill Formation · 1SLM · 1Speaker Generation · 2Speech Processing · 1Speech Recognition · 1Speech Synthesis · 33Speech-to-Speech · 1Spontaneous Speech · 1SR · 1Stock Price Prediction · 2Style Embedding · 2Survey · 2Test-Time Training · 1Text to Image · 1Time Series · 5Transformer · 5TTS · 23VAE · 2Vocoder · 2World Model · 2Zero-Shot · 3

Agent

이번 주 AI/ML 논문 훑어보기 (26.03.16 ~ 03.22) ↗이번 주 AI/ML 논문 훑어보기 (26.03.09 ~ 03.15) ↗이번 주 AI/ML 논문 훑어보기 (26.03.02 ~ 03.08) ↗이번 주 AI/ML 논문 훑어보기 (26.02.23 ~ 03.01) ↗이번 주 AI/ML 논문 훑어보기 (26.02.09 ~ 02.15) ↗이번 주 AI/ML 논문 훑어보기 (26.02.02 ~ 02.08) ↗이번 주 AI/ML 논문 훑어보기 (26.01.26 ~ 02.01) ↗

Agentic RL

이번 주 AI/ML 논문 훑어보기 (26.03.09 ~ 03.15) ↗

AI Agent

Unit 1: Introduction to Agents (Part 1) ↗

AI Ops

이번 주 AI/ML 논문 훑어보기 (26.02.23 ~ 03.01) ↗

Anti Spoofing

RawBoost: A Raw Data Boosting and Augmentation Method applied to Automatic Speaker Verification Anti-Spoofing ↗Aasist: Audio Anti-Spoofing Using Integrated Spectro-Temporal Graph Attention Networks ↗

ASR

End-to-end Music-mixed Speech Recognition ↗Deep Representation Learning in Speech Processing: Challenges, Recent Advances, and Future Trends ↗

Audio Augmentation

RawBoost: A Raw Data Boosting and Augmentation Method applied to Automatic Speaker Verification Anti-Spoofing ↗

Audio Compression

High Fidelity Neural Audio Compression ↗SoundStream: An End-to-End Neural Audio Codec ↗

Audio Deepfake Detection

RawBoost: A Raw Data Boosting and Augmentation Method applied to Automatic Speaker Verification Anti-Spoofing ↗Aasist: Audio Anti-Spoofing Using Integrated Spectro-Temporal Graph Attention Networks ↗

Audio Synthesis

Adversarial audio synthesis ↗

Auto Harness

이번 주 AI/ML 논문 훑어보기 (26.03.09 ~ 03.15) ↗

BERT

Unified Mandarin TTS Front-end Based on Distilled BERT Model ↗

Byte representation

Multilingual grapheme-to-phoneme conversion with byte representation ↗

Code Switching

Speak Foreign Languages with Your Own Voice: Cross-Lingual Neural Codec Language Modeling ↗

Codec

High Fidelity Neural Audio Compression ↗SoundStream: An End-to-End Neural Audio Codec ↗

Continual Learning

이번 주 AI/ML 논문 훑어보기 (26.02.23 ~ 03.01) ↗

Contrastive Learning

Label Embedding for Chinese Grapheme-to-Phoneme Conversion ↗

Cross-Lingual TTS

Speak Foreign Languages with Your Own Voice: Cross-Lingual Neural Codec Language Modeling ↗

Custom Voice

AdaSpeech 3: Adaptive text to speech for spontaneous style ↗AdaSpeech 2: Adaptive text to speech with untranscribed data ↗AdaSpeech: Adaptive text to speech for custom voice ↗

Diffusion

Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speech ↗

DLM

이번 주 AI/ML 논문 훑어보기 (26.03.09 ~ 03.15) ↗

Duration

Should you use a probabilistic duration model in TTS? Probably! Especially for spontaneous speech ↗

Dynamic Time Warping

Shape and Time Distortion Loss for Training Deep Time Series Forecasting Models ↗

Finance

Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting ↗Stock Market Forecasting Using Computational Intelligence: A Survey ↗

Financial Engineering

처음 만나는 금융공학 ↗

Flow

Wave-Tacotron: Spectrogram-free end-to-end text-to-speech synthesis ↗Glow-TTS: A generative flow for text-to-speech via monotonic alignment search ↗

Flow Matching

Should you use a probabilistic duration model in TTS? Probably! Especially for spontaneous speech ↗

Full Stack Agent

이번 주 AI/ML 논문 훑어보기 (26.02.09 ~ 02.15) ↗

G2P

NN-KOG2P: A Novel Grapheme-To-Phoneme Model for Korean Language ↗Label Embedding for Chinese Grapheme-to-Phoneme Conversion ↗Neural Lexicon Reader: Reduce Pronunciation Errors in End-to-end TTS by Leveraging External Textual Knowledge ↗Multilingual grapheme-to-phoneme conversion with byte representation ↗Unified Mandarin TTS Front-end Based on Distilled BERT Model ↗A bi-directional LSTM approach for polyphone disambiguation in mandarin chinese ↗Sequence-to-sequence neural net models for grapheme-to-phoneme conversion ↗Grapheme-to-phoneme conversion using Long Short-Term Memory recurrent neural networks ↗

GAN

HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis ↗Melgan: Generative adversarial networks for conditional waveform synthesis ↗Parallel WaveGAN: A fast waveform generation model based on generative adversarial networks with multi-resolution spectrogram ↗Adversarial audio synthesis ↗Adversarial sparse transformer for time series forecasting ↗

GNN

Aasist: Audio Anti-Spoofing Using Integrated Spectro-Temporal Graph Attention Networks ↗

HACRL

이번 주 AI/ML 논문 훑어보기 (26.03.02 ~ 03.08) ↗

Hugging Face

Unit 1: Introduction to Agents (Part 1) ↗

Lexicon

Neural Lexicon Reader: Reduce Pronunciation Errors in End-to-end TTS by Leveraging External Textual Knowledge ↗

LLM

이번 주 AI/ML 논문 훑어보기 (26.03.02 ~ 03.08) ↗이번 주 AI/ML 논문 훑어보기 (26.02.23 ~ 03.01) ↗이번 주 AI/ML 논문 훑어보기 (26.02.09 ~ 02.15) ↗이번 주 AI/ML 논문 훑어보기 (26.01.26 ~ 02.01) ↗Unit 1: Introduction to Agents (Part 1) ↗

Long Sequence

Informer: Beyond Efficient Transformer for Long Sequence Time-Series ↗

LoRA

이번 주 AI/ML 논문 훑어보기 (26.03.09 ~ 03.15) ↗

Loss Function

Shape and Time Distortion Loss for Training Deep Time Series Forecasting Models ↗

LSTM

A bi-directional LSTM approach for polyphone disambiguation in mandarin chinese ↗Grapheme-to-phoneme conversion using Long Short-Term Memory recurrent neural networks ↗

Mamba3

이번 주 AI/ML 논문 훑어보기 (26.03.16 ~ 03.22) ↗

Meta Learning

Meta-StyleSpeech: Multi-Speaker Adaptive Text-to-Speech Generation ↗

MetaClaw

이번 주 AI/ML 논문 훑어보기 (26.03.16 ~ 03.22) ↗

MoE

이번 주 AI/ML 논문 훑어보기 (26.03.02 ~ 03.08) ↗

MOS

Generalization Ability of MOS Prediction Networks ↗LDNet: Unified Listener Dependent Modeling in MOS Prediction for Synthetic Speech ↗MBNet: MOS Prediction for Synthesized Speech with Mean-Bias Network ↗

Multi Speaker

Multi-Speaker Emotional Text-to-Speech Synthesizer ↗

Natural Language Guidance

Natural language guidance of high-fidelity text-to-speech with synthetic annotations ↗

Neural Codec Language Model

VALL-E 2: Neural Codec Language Models are Human Parity Zero-Shot Text to Speech Synthesizers ↗VALL-E R: Robust and Efficient Zero-Shot Text-to-Speech Synthesis via Monotonic Alignment ↗Speak Foreign Languages with Your Own Voice: Cross-Lingual Neural Codec Language Modeling ↗Neural Codec Langauge Models are Zero-Shot Text to Speech Synthesizers ↗

Neural Vocoder

HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis ↗

On-Device AI

이번 주 AI/ML 논문 훑어보기 (26.02.09 ~ 02.15) ↗

Online Course

Unit 1: Introduction to Agents (Part 1) ↗

OpenSeeker

이번 주 AI/ML 논문 훑어보기 (26.03.16 ~ 03.22) ↗

Positional Embedding

이번 주 AI/ML 논문 훑어보기 (26.01.26 ~ 02.01) ↗

Privacy

이번 주 AI/ML 논문 훑어보기 (26.02.23 ~ 03.01) ↗

Prompt TTS

Natural language guidance of high-fidelity text-to-speech with synthetic annotations ↗

Quantization

이번 주 AI/ML 논문 훑어보기 (26.02.09 ~ 02.15) ↗

Research Automation

이번 주 AI/ML 논문 훑어보기 (26.02.02 ~ 02.08) ↗

RL

이번 주 AI/ML 논문 훑어보기 (26.03.16 ~ 03.22) ↗

Self Evolution

이번 주 AI/ML 논문 훑어보기 (26.02.02 ~ 02.08) ↗

Seq2Seq

Sequence-to-sequence neural net models for grapheme-to-phoneme conversion ↗

SER

Deep Representation Learning in Speech Processing: Challenges, Recent Advances, and Future Trends ↗

Skill Formation

이번 주 AI/ML 논문 훑어보기 (26.02.02 ~ 02.08) ↗

SLM

이번 주 AI/ML 논문 훑어보기 (26.01.26 ~ 02.01) ↗

Speaker Generation

Imaginary Voice: Face-styled Diffusion Model for Text-to-Speech ↗VoiceMe: Personalized voice generation in TTS ↗

Speech Processing

Deep Representation Learning in Speech Processing: Challenges, Recent Advances, and Future Trends ↗

Speech Recognition

Almost Unsupervised Text to Speech and Automatic Speech Recognition ↗

Speech Synthesis

Imaginary Voice: Face-styled Diffusion Model for Text-to-Speech ↗VoiceMe: Personalized voice generation in TTS ↗NN-KOG2P: A Novel Grapheme-To-Phoneme Model for Korean Language ↗Grad-TTS: A Diffusion Probabilistic Model for Text-to-Speech ↗Label Embedding for Chinese Grapheme-to-Phoneme Conversion ↗Almost Unsupervised Text to Speech and Automatic Speech Recognition ↗Neural Lexicon Reader: Reduce Pronunciation Errors in End-to-end TTS by Leveraging External Textual Knowledge ↗Multilingual grapheme-to-phoneme conversion with byte representation ↗Unified Mandarin TTS Front-end Based on Distilled BERT Model ↗A bi-directional LSTM approach for polyphone disambiguation in mandarin chinese ↗Sequence-to-sequence neural net models for grapheme-to-phoneme conversion ↗Grapheme-to-phoneme conversion using Long Short-Term Memory recurrent neural networks ↗Multi-Speaker Emotional Text-to-Speech Synthesizer ↗Style Tokens: Unsupervised style modeling, control and transfer in end-to-end speech synthesis ↗Generalization Ability of MOS Prediction Networks ↗LDNet: Unified Listener Dependent Modeling in MOS Prediction for Synthetic Speech ↗MBNet: MOS Prediction for Synthesized Speech with Mean-Bias Network ↗FastPitchFormant: Source-Filter Based Decomposed Modeling for Speech Synthesis ↗FastPitch: Parallel text-to-speech with pitch prediction ↗Meta-StyleSpeech: Multi-Speaker Adaptive Text-to-Speech Generation ↗HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis ↗Parallel Tacotron 2: A Non-Autoregressive Neural TTS Model with Differentiable Duration Modeling ↗Parallel Tacotron: Non-autoregressive and controllable TTS ↗AdaSpeech 3: Adaptive text to speech for spontaneous style ↗AdaSpeech 2: Adaptive text to speech with untranscribed data ↗AdaSpeech: Adaptive text to speech for custom voice ↗Wave-Tacotron: Spectrogram-free end-to-end text-to-speech synthesis ↗Glow-TTS: A generative flow for text-to-speech via monotonic alignment search ↗Melgan: Generative adversarial networks for conditional waveform synthesis ↗Parallel WaveGAN: A fast waveform generation model based on generative adversarial networks with multi-resolution spectrogram ↗FastSpeech 2: Fast and high-quality end-to-end text to speech ↗FastSpeech: Fast, robust and controllable text to speech ↗All about Tacotron 1,2 ↗

Speech-to-Speech

Speak Foreign Languages with Your Own Voice: Cross-Lingual Neural Codec Language Modeling ↗

Spontaneous Speech

Should you use a probabilistic duration model in TTS? Probably! Especially for spontaneous speech ↗

SR

Deep Representation Learning in Speech Processing: Challenges, Recent Advances, and Future Trends ↗

Stock Price Prediction

Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting ↗Stock Market Forecasting Using Computational Intelligence: A Survey ↗

Style Embedding

Style Tokens: Unsupervised style modeling, control and transfer in end-to-end speech synthesis ↗Meta-StyleSpeech: Multi-Speaker Adaptive Text-to-Speech Generation ↗

Survey

Stock Market Forecasting Using Computational Intelligence: A Survey ↗Deep Representation Learning in Speech Processing: Challenges, Recent Advances, and Future Trends ↗

Test-Time Training

이번 주 AI/ML 논문 훑어보기 (26.01.26 ~ 02.01) ↗

Text to Image

이번 주 AI/ML 논문 훑어보기 (26.02.02 ~ 02.08) ↗

Time Series

Informer: Beyond Efficient Transformer for Long Sequence Time-Series ↗Shape and Time Distortion Loss for Training Deep Time Series Forecasting Models ↗Adversarial sparse transformer for time series forecasting ↗N-BEATS: Neural Basis Expansion Analysis for Interpretable Time Series Forecasting ↗Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting ↗

Transformer

FastSpeech 2: Fast and high-quality end-to-end text to speech ↗FastSpeech: Fast, robust and controllable text to speech ↗Informer: Beyond Efficient Transformer for Long Sequence Time-Series ↗Adversarial sparse transformer for time series forecasting ↗Temporal Fusion Transformers for Interpretable Multi-horizon Time Series Forecasting ↗

TTS

VALL-E 2: Neural Codec Language Models are Human Parity Zero-Shot Text to Speech Synthesizers ↗VALL-E R: Robust and Efficient Zero-Shot Text-to-Speech Synthesis via Monotonic Alignment ↗Should you use a probabilistic duration model in TTS? Probably! Especially for spontaneous speech ↗Neural Codec Langauge Models are Zero-Shot Text to Speech Synthesizers ↗Multi-Speaker Emotional Text-to-Speech Synthesizer ↗Style Tokens: Unsupervised style modeling, control and transfer in end-to-end speech synthesis ↗Generalization Ability of MOS Prediction Networks ↗LDNet: Unified Listener Dependent Modeling in MOS Prediction for Synthetic Speech ↗MBNet: MOS Prediction for Synthesized Speech with Mean-Bias Network ↗FastPitchFormant: Source-Filter Based Decomposed Modeling for Speech Synthesis ↗FastPitch: Parallel text-to-speech with pitch prediction ↗Meta-StyleSpeech: Multi-Speaker Adaptive Text-to-Speech Generation ↗HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis ↗Parallel Tacotron 2: A Non-Autoregressive Neural TTS Model with Differentiable Duration Modeling ↗Parallel Tacotron: Non-autoregressive and controllable TTS ↗AdaSpeech 3: Adaptive text to speech for spontaneous style ↗AdaSpeech 2: Adaptive text to speech with untranscribed data ↗AdaSpeech: Adaptive text to speech for custom voice ↗Wave-Tacotron: Spectrogram-free end-to-end text-to-speech synthesis ↗Glow-TTS: A generative flow for text-to-speech via monotonic alignment search ↗FastSpeech 2: Fast and high-quality end-to-end text to speech ↗FastSpeech: Fast, robust and controllable text to speech ↗All about Tacotron 1,2 ↗

VAE

Parallel Tacotron 2: A Non-Autoregressive Neural TTS Model with Differentiable Duration Modeling ↗Parallel Tacotron: Non-autoregressive and controllable TTS ↗

Vocoder

Melgan: Generative adversarial networks for conditional waveform synthesis ↗Parallel WaveGAN: A fast waveform generation model based on generative adversarial networks with multi-resolution spectrogram ↗

World Model

이번 주 AI/ML 논문 훑어보기 (26.03.02 ~ 03.08) ↗이번 주 AI/ML 논문 훑어보기 (26.02.09 ~ 02.15) ↗

Zero-Shot

VALL-E 2: Neural Codec Language Models are Human Parity Zero-Shot Text to Speech Synthesizers ↗VALL-E R: Robust and Efficient Zero-Shot Text-to-Speech Synthesis via Monotonic Alignment ↗Neural Codec Langauge Models are Zero-Shot Text to Speech Synthesizers ↗