Search Results for author: Kyungjae Lee

Found 28 papers, 8 papers with code

On Monotonic Aggregation for Open-domain QA

1 code implementation8 Aug 2023 Sang-eun Han, Yeonseok Jeong, Seung-won Hwang, Kyungjae Lee

Our experiments show that our framework not only ensures monotonicity, but also outperforms state-of-the-art multi-source QA methods on Natural Questions.

Language Modelling Natural Questions +4

When to Read Documents or QA History: On Unified and Selective Open-domain QA

no code implementations7 Jun 2023 Kyungjae Lee, Sang-eun Han, Seung-won Hwang, Moontae Lee

This paper studies the problem of open-domain question answering, with the aim of answering a diverse range of questions leveraging knowledge resources.

Natural Questions Open-Domain Question Answering +2

Revisiting Dense Retrieval with Unanswerable Counterfactuals

no code implementations6 Apr 2023 Yongho Song, Dahyun Lee, Kyungjae Lee, Jinyeong Yeo

The retriever-reader framework is popular for open-domain question answering (ODQA), where a retriever samples for the reader a set of relevant candidate passages from a large corpus.

Contrastive Learning Open-Domain Question Answering +3

Exploring the Benefits of Training Expert Language Models over Instruction Tuning

1 code implementation7 Feb 2023 Joel Jang, Seungone Kim, Seonghyeon Ye, Doyoung Kim, Lajanugen Logeswaran, Moontae Lee, Kyungjae Lee, Minjoon Seo

Recently, Language Models (LMs) instruction-tuned on multiple tasks, also known as multitask-prompted fine-tuning (MT), have shown the capability to generalize to unseen tasks.

Efficient Trust Region-Based Safe Reinforcement Learning with Low-Bias Distributional Actor-Critic

no code implementations26 Jan 2023 Dohyeong Kim, Kyungjae Lee, Songhwai Oh

In this paper, we develop a safe distributional RL method based on the trust region method, which can satisfy constraints consistently.

Reinforcement Learning (RL) Safe Reinforcement Learning

Look Around for Anomalies: Weakly-Supervised Anomaly Detection via Context-Motion Relational Learning

no code implementations CVPR 2023 MyeongAh Cho, Minjung Kim, Sangwon Hwang, Chaewon Park, Kyungjae Lee, Sangyoun Lee

Furthermore, as the relationship between context and motion is important in order to identify the anomalies in complex and diverse scenes, we propose a Context--Motion Interrelation Module (CoMo), which models the relationship between the appearance of the surroundings and motion, rather than utilizing only temporal dependencies or motion information.

Relational Reasoning Supervised Anomaly Detection +2

Plug-and-Play Adaptation for Continuously-updated QA

no code implementations Findings (ACL) 2022 Kyungjae Lee, Wookje Han, Seung-won Hwang, Hwaran Lee, Joonsuk Park, Sang-Woo Lee

To this end, we first propose a novel task--Continuously-updated QA (CuQA)--in which multiple large-scale updates are made to LMs, and the performance is measured with respect to the success in adding and updating knowledge while retaining existing knowledge.

Domain Generalization by Mutual-Information Regularization with Pre-trained Models

1 code implementation21 Mar 2022 Junbum Cha, Kyungjae Lee, Sungrae Park, Sanghyuk Chun

Domain generalization (DG) aims to learn a generalized model to an unseen target domain using only limited source domains.

Domain Generalization

Neural Markov Controlled SDE: Stochastic Optimization for Continuous-Time Data

no code implementations ICLR 2022 Sung Woo Park, Kyungjae Lee, Junseok Kwon

We propose a novel probabilistic framework for modeling stochastic dynamics with the rigorous use of stochastic optimal control theory.

Stochastic Optimization Time Series +1

Semi-Autonomous Teleoperation via Learning Non-Prehensile Manipulation Skills

no code implementations27 Sep 2021 Sangbeom Park, Yoonbyung Chai, Sunghyun Park, Jeongeun Park, Kyungjae Lee, Sungjoon Choi

In this paper, we present a semi-autonomous teleoperation framework for a pick-and-place task using an RGB-D sensor.

Query Generation for Multimodal Documents

no code implementations EACL 2021 Kyungho Kim, Kyungjae Lee, Seung-won Hwang, Young-In Song, SeungWook Lee

This paper studies the problem of generatinglikely queries for multimodal documents withimages.


Optimal Algorithms for Stochastic Multi-Armed Bandits with Heavy Tailed Rewards

no code implementations NeurIPS 2020 Kyungjae Lee, Hongjun Yang, Sungbin Lim, Songhwai Oh

In simulation, the proposed estimator shows favorable performance compared to existing robust estimators for various $p$ values and, for MAB problems, the proposed perturbation strategy outperforms existing exploration methods.

Multi-Armed Bandits

Relational Deep Feature Learning for Heterogeneous Face Recognition

no code implementations2 Mar 2020 MyeongAh Cho, Taeoh Kim, Ig-Jae Kim, Kyungjae Lee, Sangyoun Lee

Due to the lack of databases, HFR methods usually exploit the pre-trained features on a large-scale visual database that contain general facial information.

Face Recognition Heterogeneous Face Recognition

Categorical Metadata Representation for Customized Text Classification

2 code implementations TACL 2019 Jihyeok Kim, Reinald Kim Amplayo, Kyungjae Lee, Sua Sung, Minji Seo, Seung-won Hwang

The performance of text classification has improved tremendously using intelligently engineered neural-based models, especially those injecting categorical metadata as additional information, e. g., using user/product information for sentiment classification.

Ranked #4 on Sentiment Analysis on User and product information (Yelp 2013 (Acc) metric)

General Classification Sentiment Analysis +4

Tsallis Reinforcement Learning: A Unified Framework for Maximum Entropy Reinforcement Learning

no code implementations31 Jan 2019 Kyungjae Lee, Sungyub Kim, Sungbin Lim, Sungjoon Choi, Songhwai Oh

By controlling the entropic index, we can generate various types of entropy, including the SG entropy, and a different entropy results in a different class of the optimal policy in Tsallis MDPs.

reinforcement-learning Reinforcement Learning (RL)

ChoiceNet: Robust Learning by Revealing Output Correlations

no code implementations27 Sep 2018 Sungjoon Choi, Sanghoon Hong, Kyungjae Lee, Sungbin Lim

To this end, we present a novel framework referred to here as ChoiceNet that can robustly infer the target distribution in the presence of inconsistent data.


Maximum Causal Tsallis Entropy Imitation Learning

no code implementations NeurIPS 2018 Kyungjae Lee, Sungjoon Choi, Songhwai Oh

Third, we propose a maximum causal Tsallis entropy imitation learning (MCTEIL) algorithm with a sparse mixture density network (sparse MDN) by modeling mixture weights using a sparsemax distribution.

Imitation Learning

Task Agnostic Robust Learning on Corrupt Outputs by Correlation-Guided Mixture Density Networks

1 code implementation CVPR 2020 Sungjoon Choi, Sanghoon Hong, Kyungjae Lee, Sungbin Lim

In this paper, we focus on weakly supervised learning with noisy training data for both classification and regression problems. We assume that the training outputs are collected from a mixture of a target and correlated noise distributions. Our proposed method simultaneously estimates the target distribution and the quality of each data which is defined as the correlation between the target and data generating distributions. The cornerstone of the proposed method is a Cholesky Block that enables modeling dependencies among mixture distributions in a differentiable manner where we maintain the distribution over the network weights. We first provide illustrative examples in both regression and classification tasks to show the effectiveness of the proposed method. Then, the proposed method is extensively evaluated in a number of experiments where we show that it constantly shows comparable or superior performances compared to existing baseline methods in the handling of noisy data.

Autonomous Driving General Classification +2

Uncertainty-Aware Learning from Demonstration using Mixture Density Networks with Sampling-Free Variance Modeling

1 code implementation3 Sep 2017 Sungjoon Choi, Kyungjae Lee, Sungbin Lim, Songhwai Oh

The proposed uncertainty-aware learning from demonstration method outperforms other compared methods in terms of safety using a complex real-world driving dataset.

Autonomous Driving

Density Matching Reward Learning

no code implementations12 Aug 2016 Sungjoon Choi, Kyungjae Lee, Andy Park, Songhwai Oh

The performance of KDMRL is extensively evaluated in two sets of experiments: grid world and track driving experiments.

Autonomous Navigation

Cannot find the paper you are looking for? You can Submit a new open access paper.