Taehyeon Kim
YOU?
Author Swipe
View article: Few-Shot Demonstration-Driven Task Coordination and Trajectory Execution for Multi-Robot Systems
Few-Shot Demonstration-Driven Task Coordination and Trajectory Execution for Multi-Robot Systems Open
In this paper, we propose a novel few-shot learning framework for multi-robot systems that integrate both spatial and temporal elements: Few-Shot Demonstration-Driven Task Coordination and Trajectory Execution (DDACE). Our approach leverag…
View article: Guiding Reasoning in Small Language Models with LLM Assistance
Guiding Reasoning in Small Language Models with LLM Assistance Open
The limited reasoning capabilities of small language models (SLMs) cast doubt on their suitability for tasks demanding deep, multi-step logical deduction. This paper introduces a framework called Small Reasons, Large Hints (SMART), which s…
View article: 3D Digital Human Generation from a Single Image Using Generative AI with Real-Time Motion Synchronization
3D Digital Human Generation from a Single Image Using Generative AI with Real-Time Motion Synchronization Open
The generation of 3D digital humans has traditionally relied on multi-view imaging systems and large-scale datasets, posing challenges in cost, accessibility, and real-time applicability. To overcome these limitations, this study presents …
View article: A BERT-based Denoising Framework for Inter-Company Similarity Analysis Using 10-K Reports
A BERT-based Denoising Framework for Inter-Company Similarity Analysis Using 10-K Reports Open
View article: REBEL: Rule-based and Experience-enhanced Learning with LLMs for Initial Task Allocation in Multi-Human Multi-Robot Teaming
REBEL: Rule-based and Experience-enhanced Learning with LLMs for Initial Task Allocation in Multi-Human Multi-Robot Teaming Open
Multi-human multi-robot teams are increasingly recognized for their efficiency in executing large-scale, complex tasks by integrating heterogeneous yet potentially synergistic humans and robots. However, this inherent heterogeneity present…
View article: Adaptive Task Allocation in Multi-Human Multi-Robot Teams under Team Heterogeneity and Dynamic Information Uncertainty
Adaptive Task Allocation in Multi-Human Multi-Robot Teams under Team Heterogeneity and Dynamic Information Uncertainty Open
Task allocation in multi-human multi-robot (MH-MR) teams presents significant challenges due to the inherent heterogeneity of team members, the dynamics of task execution, and the information uncertainty of operational states. Existing app…
View article: PrefMMT: Modeling Human Preferences in Preference-based Reinforcement Learning with Multimodal Transformers
PrefMMT: Modeling Human Preferences in Preference-based Reinforcement Learning with Multimodal Transformers Open
Preference-based reinforcement learning (PbRL) shows promise in aligning robot behaviors with human preferences, but its success depends heavily on the accurate modeling of human preferences through reward models. Most methods adopt Markov…
View article: Towards Fast Multilingual LLM Inference: Speculative Decoding and Specialized Drafters
Towards Fast Multilingual LLM Inference: Speculative Decoding and Specialized Drafters Open
Large language models (LLMs) have revolutionized natural language processing and broadened their applicability across diverse commercial applications. However, the deployment of these models is constrained by high inference time in multili…
View article: Block Transformer: Global-to-Local Language Modeling for Fast Inference
Block Transformer: Global-to-Local Language Modeling for Fast Inference Open
We introduce the Block Transformer which adopts hierarchical global-to-local modeling to autoregressive transformers to mitigate the inference bottlenecks associated with self-attention. Self-attention requires the key-value (KV) cache of …
View article: Exploring and Improving Drafts in Blockwise Parallel Decoding
Exploring and Improving Drafts in Blockwise Parallel Decoding Open
Despite the remarkable strides made by autoregressive language models, their potential is often hampered by the slow inference speeds inherent in sequential token generation. Blockwise parallel decoding (BPD) was proposed by Stern et al. a…
View article: Leveraging Normalization Layer in Adapters with Progressive Learning and Adaptive Distillation for Cross-Domain Few-Shot Learning
Leveraging Normalization Layer in Adapters with Progressive Learning and Adaptive Distillation for Cross-Domain Few-Shot Learning Open
Cross-domain few-shot learning presents a formidable challenge, as models must be trained on base classes and then tested on novel classes from various domains with only a few samples at hand. While prior approaches have primarily focused …
View article: Semantic Layering in Room Segmentation via LLMs
Semantic Layering in Room Segmentation via LLMs Open
In this paper, we introduce Semantic Layering in Room Segmentation via LLMs (SeLRoS), an advanced method for semantic room segmentation by integrating Large Language Models (LLMs) with traditional 2D map-based segmentation. Unlike previous…
View article: Hypernetwork-Driven Model Fusion for Federated Domain Generalization
Hypernetwork-Driven Model Fusion for Federated Domain Generalization Open
Federated Learning (FL) faces significant challenges with domain shifts in heterogeneous data, degrading performance. Traditional domain generalization aims to learn domain-invariant features, but the federated nature of model averaging of…
View article: Revisiting Early-Learning Regularization When Federated Learning Meets Noisy Labels
Revisiting Early-Learning Regularization When Federated Learning Meets Noisy Labels Open
In the evolving landscape of federated learning (FL), addressing label noise presents unique challenges due to the decentralized and diverse nature of data collection across clients. Traditional centralized learning approaches to mitigate …
View article: Polyphonic Piano Music Transcription System Exploiting Mutual Correlations of Different Musical Note States
Polyphonic Piano Music Transcription System Exploiting Mutual Correlations of Different Musical Note States Open
Generally, polyphonic piano music transcription systems are designed to estimate and determine pitch activities along with various note states for each audio frame. While the music transcription system has multiple uses in the Music Inform…
View article: Leveraging Normalization Layer in Adapters With Progressive Learning and Adaptive Distillation for Cross-Domain Few-Shot Learning
Leveraging Normalization Layer in Adapters With Progressive Learning and Adaptive Distillation for Cross-Domain Few-Shot Learning Open
Cross-domain few-shot learning presents a formidable challenge, as models must be trained on base classes and then tested on novel classes from various domains with only a few samples at hand. While prior approaches have primarily focused …
View article: Bridging Real and Virtual: Human Digital Twin Strategies for Workspaces
Bridging Real and Virtual: Human Digital Twin Strategies for Workspaces Open
The Human Digital Twin (HDT) is a pivotal element in smart manufacturing systems geared towards Industry 5.0. HDT represents a digital manifestation of humans, aiming to innovate the integration between humans and systems by directly linki…
View article: WiFi's Unspoken Tales: Deep Neural Network Decodes Human Behavior from Channel State Information
WiFi's Unspoken Tales: Deep Neural Network Decodes Human Behavior from Channel State Information Open
WiFi Channel State Information (CSI) represents the characteristics of wireless channels in wireless networks. WiFi CSI plays a pivotal role in wireless communications, primarily due to the variability of channel characteristics across tim…
View article: Instructive Decoding: Instruction-Tuned Large Language Models are Self-Refiner from Noisy Instructions
Instructive Decoding: Instruction-Tuned Large Language Models are Self-Refiner from Noisy Instructions Open
While instruction-tuned language models have demonstrated impressive zero-shot generalization, these models often struggle to generate accurate responses when faced with instructions that fall outside their training set. This paper present…
View article: Navigating Data Heterogeneity in Federated Learning A Semi-Supervised Federated Object Detection
Navigating Data Heterogeneity in Federated Learning A Semi-Supervised Federated Object Detection Open
Federated Learning (FL) has emerged as a potent framework for training models across distributed data sources while maintaining data privacy. Nevertheless, it faces challenges with limited high-quality labels and non-IID client data, parti…
View article: Traffic Sign Recognition Based on Bayesian Angular Margin Loss for an Autonomous Vehicle
Traffic Sign Recognition Based on Bayesian Angular Margin Loss for an Autonomous Vehicle Open
Traffic sign recognition is a pivotal technology in the advancement of autonomous vehicles as it is critical for adhering to country- or region-specific traffic regulations. Defined as an image classification problem in computer vision, tr…
View article: An Indoor Multi-Environment Sensor System Based on Intelligent Edge Computing
An Indoor Multi-Environment Sensor System Based on Intelligent Edge Computing Open
Monitoring and predicting the environment in an indoor space plays an important role in securing big data and detecting abnormal conditions in the industrial environment and living space. This study proposes an indoor multi-environment sen…
View article: Region-Conditioned Orthogonal 3D U-Net for Weather4Cast Competition
Region-Conditioned Orthogonal 3D U-Net for Weather4Cast Competition Open
The Weather4Cast competition (hosted by NeurIPS 2022) required competitors to predict super-resolution rain movies in various regions of Europe when low-resolution satellite contexts covering wider regions are given. In this paper, we show…
View article: Gradient-Applied Weighted Loss for Details of 3D Shape in Single-View Reconstruction
Gradient-Applied Weighted Loss for Details of 3D Shape in Single-View Reconstruction Open
There has been considerable research on reconstructing 3D shapes from single-view images; however, preserving the detailed information of the input image remains difficult. In this paper, we propose the application of a gradient map to tra…
View article: Benchmark Dataset for Precipitation Forecasting by Post-Processing the Numerical Weather Prediction
Benchmark Dataset for Precipitation Forecasting by Post-Processing the Numerical Weather Prediction Open
Precipitation forecasting is an important scientific challenge that has wide-reaching impacts on society. Historically, this challenge has been tackled using numerical weather prediction (NWP) models, grounded on physics-based simulations.…
View article: Revisiting Architecture-aware Knowledge Distillation: Smaller Models and Faster Search
Revisiting Architecture-aware Knowledge Distillation: Smaller Models and Faster Search Open
Knowledge Distillation (KD) has recently emerged as a popular method for compressing neural networks. In recent studies, generalized distillation methods that find parameters and architectures of student models at the same time have been p…
View article: Supernet Training for Federated Image Classification under System Heterogeneity
Supernet Training for Federated Image Classification under System Heterogeneity Open
Efficient deployment of deep neural networks across many devices and resource constraints, particularly on edge devices, is one of the most challenging problems in the presence of data-privacy preservation issues. Conventional approaches h…
View article: A Survey of Supernet Optimization and its Applications: Spatial and Temporal Optimization for Neural Architecture Search
A Survey of Supernet Optimization and its Applications: Spatial and Temporal Optimization for Neural Architecture Search Open
This survey focuses on categorizing and evaluating the methods of supernet optimization in the field of Neural Architecture Search (NAS). Supernet optimization involves training a single, over-parameterized network that encompasses the sea…
View article: Mold into a Graph: Efficient Bayesian Optimization over Mixed-Spaces
Mold into a Graph: Efficient Bayesian Optimization over Mixed-Spaces Open
Real-world optimization problems are generally not just black-box problems, but also involve mixed types of inputs in which discrete and continuous variables coexist. Such mixed-space optimization possesses the primary challenge of modelin…
View article: Automated Filter Pruning Based on High-Dimensional Bayesian Optimization
Automated Filter Pruning Based on High-Dimensional Bayesian Optimization Open
Filter pruning is necessary to efficiently deploy convolutional neural networks on edge devices that have limited computational resources and power budgets. With conventional filter pruning techniques, the same pruning rate is manually spe…