Our paper Expression Domain Translation Network for Cross-domain Head Reenactment (Kang et al.) has been accepted to ICASSP 2024.

Our papers MagiCapture: High-Resolution Multi-Concept Portrait Customization (Hyung et al.), iDet3D: Towards Efficient Interactive Object Detection for LiDAR Point Clouds (Choi et al.), and When Model Meets New Normals: Test-time Adaptation for Unsupervised Time-series Anomaly Detection (Kim et al.) have been accepted to AAAI 2024.

Our papers LensNeRF: Rethinking Volume Rendering based on Thin-Lens Camera Model (Kim et al.) and Slice and Conquer: A Planar-to-3D Framework for Efficient Interactive Segmentation of Volumetric Images (Cho et al.) have been accepted to WACV 2024.

Our papers SimCKP: Simple Contrastive Learning of Keyphrase Representations (Choi et al.), AniEE: A Dataset of Animal Experimental Literature for Event Extraction (Kim et al.), and Towards Formality-Aware Neural Machine Translation by Leveraging Context Information (Kim et al.) have been accepted to Findings of EMNLP 2023.

Our papers Learning to Discover Skills through Guidance (Kim et al.) and PLASTIC: Improving Input and Label Plasticity for Sample Efficient Reinforcement Learning (Lee et al.) have been accepted to NeurIPS 2023.

Our paper FedGeo: Privacy-Preserving User Next Location Prediction with Federated Learning has been accepted to SIGSPATIAL 2023.

Our papers PRiSM: Enhancing Low-Resource Document-Level Relation Extraction with Relation-Aware Score Calibration (Choi et al.) and Learning to Diversify Neural Text Generation via Degenerative Model (Hong et al.) have been accepted to Findings of IJCNLP-AACL 2023.

Our papers ST-RAP: A Spatio-Temporal Framework for Real Estate Appraisal (Lee et al.) and Cracking the Code of Negative Transfer: A Cooperative Game Theoretic Approach for Cross-Domain Sequential Recommendation (Park et al.) have been accepted to CIKM 2023.

Our papers Shortcut-V2V: Compression Framework for Video-to-Video Translation based on Temporal Redundancy Reduction (Chung et al.), Learning to Generate Semantic Layouts for Higher Text-Image Correspondence in Text-to-Image Synthesis (Park et al.), Towards Open-Set Test-Time Adaptation Utilizing the Wisdom of Crowds in Entropy Minimization (Lee et al.), Label Shift Adapter for Test-Time Adaptation under Covariate and Label Shifts (Park et al.), FaceCLIPNeRF: Text-driven 3D Face Manipulation using Deformable Neural Radiance Fields (Hwang et al.), CAFA: Class-aware Feature Alignment for Test-time Adaptation (Jung et al.), 3D-Aware Generative Model for Improved Side-View Image Synthesis (Jo et al.) have been accepted to ICCV 2023.

Our papers Pre-training Contextual Location Embeddings in Personal Trajectories via Efficient Hierarchical Location Representations (Park et al.) and Deep Imbalanced Time-series Forecasting via Local Discrepancy Density (Park et al.) have been accepted to ECML/PKDD 2023.

Our paper Towards Physically Reliable Molecular Representation Learning (Yi et al.) has been accepted to UAI 2023 for oral presentation (3.1% acceptance rate).

Our paper HistRED: A Historical Document-Level Relation Extraction Dataset (Yang et al.) has been accepted to ACL 2023, and papers Towards Accurate Translation via Semantically Appropriate Application of Lexical Constraints (Baek et al.) and DEnsity: Open-domain Dialogue Evaluation Metric using Density Estimation (Park et al.) have been accepted to Findings of ACL 2023.

Our paper On the Importance of Feature Decorrelation for Unsupervised Representation Learning in Reinforcement Learning (Lee et al.) has been accepted to ICML 2023.

Our paper Local 3D Editing via 3D Distillation of CLIP Knowledge (Hyung et al.) has been accepted to CVPR 2023.

Our paper Perspective Projection-Based 3D CT Reconstruction from Biplanar X-rays (Kyung et al.) has been accepted to ICASSP 2023 and has won the Best Student Paper Award!

Our paper TTN: A Domain-Shift Aware Batch Normalization in Test-Time Adaptation (Lim et al.) has been accepted to ICLR 2023.

Our paper PePe: Personalized Post-editing Model Utilizing User-generated Post-edits (Lee et al.) has been accepted to Findings of EACL 2023.

Our paper Revisiting the Importance of Amplifying Bias for Debiasing (Lee et al.) has been accepted to AAAI 2023.

Our papers iColoriT: Towards Propagating Local Hint to the Right Region in Interactive Colorization by Leveraging Vision Transformer (Lee et al.), CG-NeRF: Conditional Generative Neural Radiance Fields for 3D-aware Image Synthesis (Jo et al.), Training Auxiliary Prototypical Classifiers for Explainable Anomaly Detection in Medical Image Segmentation (Cho et al.), Guiding Users to Where to Give Color Hints for Efficient Interactive Sketch Colorization via Unsupervised Region Prioritization (Cho et al.) are accepted to WACV 2023.

Our papers PNEG: Prompt-based Negative Response Generation for Dialogue Response Selection Task (Lee et al.) and Rethinking Style Transformer by Energy-based Interpretation: Adversarial Unsupervised Style Transfer using Pretrained Model (Cho et al.) are accepted to EMNLP 2022.

Our papers Mining Multi-Label Samples from Single Positive Labels (Cho et al.) and WaveBound: Dynamic Error Bounds for Stable Time Series Forecasting (Cho et al.) are accepted to NeurIPS 2022, and CEDe: A collection of expert-curated datasets with atom-level entity annotations for Optical Chemical Structure Recognition (Hormazabal et al.) is accepted to NeurIPS 2022 Datasets and Benchmarks Track.

Our paper Reweighting Strategy based on Synthetic Data Identification for Sentence Similarity (Kim et al.) is accepted to COLING 2022.

Sunghyun wins the Best Paper Award for his paper AnimeCeleb: Large-Scale Animation CelebHeads Dataset for Head Reenactment (Kim et al.) at KAIA Summer Conference 2022.

Our paper Residual Correction in Real-Time Traffic Forecasting (Kim et al.) is accepted to CIKM 2022.

Hojoon wins the Best Short Paper Honorable Mention Award for his paper Towards Validating Long-Term User Feedbacks in Interactive Recommendation Systems (Lee et al.) at SIGIR 2022.

Our papers Towards Accurate Open-Set Recognition via Background-Class Regularization (Cho & Choo), High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled Conditions (Lee et al.), Style Your Hair: Latent Optimization for Pose-Invariant Hairstyle Transfer via Local-Style-Aware Hair Alignment (Kim et al.), and AnimeCeleb: Large-Scale Animation CelebHeads Dataset for Head Reenactment (Kim et al.) are accepted to ECCV 2022.

Our paper ReFu: Refine and Fuse the Unobserved View for Detail-Preserving Single-Image 3D Human Reconstruction (Shim et al.) is accepted to ACMMM 2022.

Our paper Morphology-Aware Interactive Keypoint Estimation (Kim et al.) is accepted to MICCAI 2022.

Our paper Enemy Spotted: In-game Gun Sound Dataset for Gunshot Classification and Localization (Park et al.) is accepted to CoG 2022.

Our paper Towards Validating Long-Term User Feedbacks in Interactive Recommendation Systems (Lee et al.) is accepted to SIGIR 2022.

Our paper DASH: Visual Analytics for Debiasing Image Classification via User-Driven Synthetic Data Augmentation (Kwon et al.) is accepted to EuroVis 2022.

Congratulations to Prof. Jaegul Choo for becoming a tenured professor at KAIST! We got together to celebrate him and listened to his heartfelt speech :)

Letsur, a startup led by Gyuhyeon, wins one of the top 20 spots for Samsung's C-Lab Outside program and receives various business supports from Samsung Research. Check out the news article.

Our paper Reversible Instance Normalization for Accurate Time-Series Forecasting against Distribution Shift (Kim et al.) is accepted to ICLR 2022.

Our paper PASTA: Parallel Spatio-Temporal Attention with Spatial Auto-Correlation Gating for Fine-Grained Crowd Flows Prediction (Park et al.) is accepted to PAKDD 2022.

Our paper DraftRec: Personalized Draft Recommendation for Victory in Multi-Player Online Battle Arena Games (Lee et al.) is accepted to WWW 2022.

Hyesu, Jeonghoon, and Sanghyeon win the Best Poster Award for their papers AVoCaDo: Strategy for Adapting Vocabulary to Downstream Domain (Hong et al.) and Deep Edge-Aware Interactive Colorization against Color-Bleeding Effects (Kim et al.) at the 2022 KAIST AI Workshop!

Our papers Continuous-Time Video Generation via Learning Motion Dynamics with Neural ODE (Kim et al.), Improving Face Recognition with Large Age Gaps by Learning to Distinguish Children (Lee et al.), and HairFIT: Pose-Invariant Hairstyle Transfer via Flow-based Hair Alignment and Semantic-Region-Aware Inpainting (Chung et al.) are accepted to BMVC 2021. HairFit is selected for oral presentation (3.6% acceptance rate).

Our paper Learning Debiased Representation via Disentangled Feature Augmentation (Lee et al.) is accepted to NeurIPS 2021 for oral presentation (<1% acceptance rate).

Our paper Augmenting Imbalanced Time-series Data via Adversarial Perturbation in Latent Space (Kim et al.) is accepted to ACML 2021.

Our paper AVoCaDo: Strategy for Adapting Vocabulary to Downstream Domain (Hong et al.) is accepted to EMNLP 2021, and Novel Natural Language Summarization of Program Code via Leveraging Multiple Input Representations (Chen et al.) is accepted to Findings of EMNLP 2021.

Our papers BiaSwap: Removing Dataset Bias with Bias-Tailored Swapping Augmentation (Kim et al.), Deep Edge-Aware Interactive Colorization against Color-Bleeding Effects (Kim et al.), and Standardized Max Logit: A Simple yet Effective Approach for Identifying Unexpected Road Obstacles in Urban-Scene Segmentation (Jung et al.) are accepted to ICCV 2021. The second and third papers are selected for oral presentation (3% acceptance rate).

Our paper Knowledge Graph-based Question Answering with Electronic Health Records (Park et al.) is accepted to MLHC 2021. Check out the intro talk of the paper.

Our paper K-Hairstyle: A Large-Scale Korean Hairstyle Dataset for Virtual Hair Editing and Hairstyle Classification (Kim et al.) is accepted to ICIP 2021. Check out the project page of the paper.

Our papers Unsupervised Neural Machine Translation for Low-Resource Domains via Meta-Learning (Park et al.) and Constructing Multi-Modal Dialogue Dataset with Text-to-Image Replacement (Lee et al.) are accepted to ACL 2021, and Learning to Generate Questions by Learning to Recover Answer-containing Sentences (Back et al.) is accepted to Findings of ACL 2021.

Our paper VATUN: Visual Analytics for Testing and Understanding Convolutional Neural Networks (Park et al.) is accepted to EuroVis 2021. Check out the demo and talk of the paper.

Our paper Understanding Human-side Impact of Sampling Image Batches in Subjective Attribute Labeling (Chung et al.) is accepted to CSCW 2021.

Our paper Restoring and Mining the Records of the Joseon Dynasty via Neural Language Modeling and Machine Translation (Kang et al.) is accepted to NAACL 2021.

Our papers Not Just Compete, But Collaborate: Local Image-to-Image Translation via Cooperative Mask Prediction (Kim et al.), VITON-HD: High-Resolution Virtual Try-On via Misalignment-Aware Normalization (Choi et al.), and RobustNet: Improving Domain Generalization in Urban-Scene Segmentation via Instance Selective Whitening (Choi et al.) are accepted to CVPR 2021. RobustNet is selected for oral presentation (4.2% acceptance rate). Check out the code and demo of VITON-HD, as well as the code and talk of RobustNet.

Our paper Efficient Adversarial Audio Synthesis via Progressive Upsampling (Cho et al.) is accepted to ICASSP 2021.

Our paper Vid-ODE: Continuous-Time Video Generation with Neural Ordinary Differential Equation (Park et al.) is accepted to AAAI 2021. Check out the code.

Our paper An Empirical Experiment on Deep Learning Models for Predicting Traffic Data (Lee et al.) is accepted to ICDE 2021.

Our paper Multimodal Image Translation with Stochastic Style Representations and Mutual Information Loss (Na et al.) is accepted to BMVC 2020.

Our paper ST-GRAT: Spatio-Temporal Graph Attention Network for Traffic Forecasting (Park et al.) is accepted to CIKM 2020.

Our paper Learning De-biased Representations with Biased Representations (Bahng et al.) is accepted to ICML 2020. Check out the code and talk of the paper.

Our paper End-To-End Multi-Task Learning of Missing Value Imputation and Forecasting in Time-Series Data (Kim et al.) is accepted to ICPR 2020.

Our papers Exploring Unlabeled Faces for Novel Attribute Discovery (Bahng et al.), Reference-based Sketch Image Colorization using Augmented-Self Reference and Dense Semantic Correspondence (Lee et al.), and Cars Can’t Fly Up in the Sky: Improving Urban-Scene Segmentation via Height-driven Attention Networks (Choi et al.) are accepted to CVPR 2020. Check out the intro and code of the third paper.

Our paper NeurQuRI: Neural Question Requirement Inspector for Answerability Prediction in Machine Reading Comprehension (Back et al.) is accepted to ICLR 2020.

Our paper NL2pSQL: Generating Pseudo-SQL Queries from Under-specified Natural Language Question (Chen et al.) is accepted to EMNLP 2019.

Our paper Recommender System Using Sequential and Global Preference via Attention Mechanism and Topic Modeling (Kang et al.) is accepted to CIKM 2019.

Our paper SANVis: Visual Analytics for Understanding Self-Attention Networks (Park et al.) is accepted to IEEE VIS 2019. Check out the demo.

Our paper Learning to Focus and Track Extreme Climate Events (Kim et al.) is accepted to BMVC 2019 for spotlight presentation (6.9% acceptance rate).

Our papers Coloring With Limited Data: Few-Shot Colorization via Memory-Augmented Networks (Yoo et al.) and Image-to-Image Translation via Group-wise Deep Whitening and Coloring (Cho et al.) are accepted to CVPR 2019. The second paper is selected for oral presentation (5.5% acceptance rate). Check out the project page of the first paper and the talk of the second paper.

Our paper AILA: Attentive Interactive Labeling Assistant for Document Classification through Attention-Based Deep Neural Networks (Choi et al.) is accepted to CHI 2019. Check out the intro and full demo.

Our paper MemoReader: Large-Scale Reading Comprehension through Neural Memory Controller (Back et al.) is accepted to EMNLP 2018.

Our paper Coloring with Words: Guiding Image Colorization Through Text-based Palette Generation (Bahng et al.) is accepted to ECCV 2018. Check out the code.

Our paper StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation (Choi et al.) is accepted to CVPR 2018 for oral presentation (2.1% acceptance rate). Check out the code and video.

Our paper TopicOnTiles: Tile-Based Spatio-Temporal Event Analytics via Exclusive Topic Modeling on Social Media (Choi et al.) is accepted to CHI 2018. Check out the intro and full demo.