arXiv subjects

Seungmin Han

Publications and source records attributed to Seungmin Han.

4 recordsLinked to original sources

arXiv2026

Structured Frequency-Domain Evidence for LLM-Based Time-Series Anomaly Detection

Time-series anomalies can appear not only as pointwise deviations but also as changes in recurring temporal structure, such as shifted periodicity or localized oscillatory fluctuations. However, existing LLM-based time-series anomaly detection methods mainly expose time-domain evidence through indexed values, plots, or de-seasonalized representations, leaving spectral structure implicit. We propose an evidence-augmented zero-shot TSAD framework that preserves indexed de-seasonalized observations while adding compact frequency-domain evidence computed with the Fast Fourier Transform (FFT). The evidence is constructed at two resolutions: global frequency-domain evidence summarizes sequence-level periodic context, while local frequency-domain evidence captures time-localized spectral departures. Experiments on AnomLLM with InternVL2-LLaMA3-76B, Qwen2.5-VL-72B-Instruct, Gemini-2.5-Flash, and GPT-4o, together with evaluation on the TSB-AD-U subset, show that explicit frequency-domain evidence improves LLM-based TSAD baselines. These results suggest that frequency-domain evidence can complement indexed and de-seasonalized time-domain inputs for zero-shot LLM-based TSAD.

cs.LG↗

arXiv2025

ContextualLVLM-Agent: A Holistic Framework for Multi-Turn Visually-Grounded Dialogue and Complex Instruction Following

Despite significant advancements in Large Language Models (LLMs) and Large Vision-Language Models (LVLMs), current models still face substantial challenges in handling complex, multi-turn, and visually-grounded tasks that demand deep reasoning, sustained contextual understanding, entity tracking, and multi-step instruction following. Existing benchmarks often fall short in capturing the dynamism and intricacies of real-world multi-modal interactions, leading to issues such as context loss and visual hallucinations. To address these limitations, we introduce MMDR-Bench (Multi-Modal Dialogue Reasoning Benchmark), a novel dataset comprising 300 meticulously designed complex multi-turn dialogue scenarios, each averaging 5-7 turns and evaluated across six core dimensions including visual entity tracking and reasoning depth. Furthermore, we propose CoLVLM Agent (Contextual LVLM Agent), a holistic framework that enhances existing LVLMs with advanced reasoning and instruction following capabilities through an iterative "memory-perception-planning-execution" cycle, requiring no extensive re-training of the underlying models. Our extensive experiments on MMDR-Bench demonstrate that CoLVLM Agent consistently achieves superior performance, attaining an average human evaluation score of 4.03, notably surpassing state-of-the-art commercial models like GPT-4o (3.92) and Gemini 1.5 Pro (3.85). The framework exhibits significant advantages in reasoning depth, instruction adherence, and error suppression, and maintains robust performance over extended dialogue turns, validating the effectiveness of its modular design and iterative approach for complex multi-modal interactions.

cs.CL↗

arXiv2023

Time Series Forecasting for Air Pollution in Seoul

Accurate air pollution forecasting plays a crucial role in controlling air quality and minimizing adverse effects on human life. Among pollutants, atmospheric particulate matter (PM) is particularly significant, affecting both visibility and human health. In this study the concentration of air pollutants and comprehensive air quality index (CAI) data collected from 2015 to 2018 in Seoul, South Korea was analyzed. Using two different statistical models: error, trend, season (ETS) and autoregressive moving-average (ARIMA), measured monthly average PM2.5 concentration were used as input to forecast the monthly averaged concentration of PM2.5 12 months ahead. To evaluate the performance of the ETS model, five evaluation criteria were used: mean error (ME), root mean squared error (RMSE), mean absolute error (MAE), mean percentage error (MPE), and mean absolute percentage error (MAPE). Data collected from January 2019 to December 2019 were used for cross-validation check of ETS model. The best fitted ARIMA model was determined by examining the AICc (Akaike Information Criterion corrected) value. The results indicated that the ETS model outperforms the ARIMA model.

stat.AP↗

arXiv2019

Robust Full-FoV Depth Estimation in Tele-wide Camera System

Tele-wide camera system with different Field of View (FoV) lenses becomes very popular in recent mobile devices. Usually it is difficult to obtain full-FoV depth based on traditional stereo-matching methods. Pure Deep Neural Network (DNN) based depth estimation methods can obtain full-FoV depth, but have low robustness for scenarios which are not covered by training dataset. In this paper, to address the above problems we propose a hierarchical hourglass network for robust full-FoV depth estimation in tele-wide camera system, which combines the robustness of traditional stereo-matching methods with the accuracy of DNN. More specifically, the proposed network comprises three major modules: single image depth prediction module infers initial depth from input color image, depth propagation module propagates traditional stereo-matching tele-FoV depth to surrounding regions, and depth combination module fuses the initial depth with the propagated depth to generate final output. Each of these modules employs an hourglass model, which is a kind of encoder-decoder structure with skip connections. Experimental results compared with state-of-the-art depth estimation methods demonstrate that our method not only produces robust and better subjective depth quality on wild test images, but also obtains better quantitative results on standard datasets.

cs.CV↗