论文检索

Abinash Pegu

Professional photography face retouching requires achieving a balance between texture preservation and quality increase, a problem that conventional automated methods find difficult to effectively handle. We provide a new lightweight neural architecture that converts conventional dodge-and-burn photography methods into predictions for learnable additive blend maps. Instead of rebuilding whole images, our method uses a small U-Net that predicts pixel-level changes, allowing for exact brightness adjustments while maintaining the original skin texture. With a 6MB model that operates effectively on common hardware, the technique produces high-quality results while preserving texture fidelity, which is crucial for professional applications. Experimental validation offers significant computing advantages while demonstrating competitive performance with current methods.

Sibendu Paul, Haotian Jiang, Caren Chen

Automatically extracting engaging and high-quality humorous scenes from cinematic titles is pivotal for creating captivating video previews and snackable content, boosting user engagement on streaming platforms. Long-form cinematic titles, with their extended duration and complex narratives, challenge scene localization, while humor’s reliance on diverse modalities and its nuanced style add further complexity. This paper introduces an end-to-end system for automatically identifying and ranking humorous scenes from long-form cinematic titles, featuring shot detection, multimodal scene localization, and humor tagging optimized for cinematic content. Key innovations include a novel scene segmentation approach combining visual and textual cues, improved shot representations via guided triplet mining, and a multimodal humor tagging framework leveraging both audio and text modalities. Our system achieves an 18.3% AP improvement over state-of-the-art scene detection on the OVSD dataset and an F1 score of 0.834 for detecting humor in long text. Extensive evaluations across five cinematic titles demonstrate 87% of clips extracted by our pipeline are intended to be funny, while 98% of scenes are accurately localized. With successful generalization to trailers, these results showcase the pipeline’s potential to enhance content creation workflows, improve user engagement, and streamline snackable content generation for diverse cinematic media formats.

Unmesh Padalkar

In daily fantasy sports (DFS), match participation is highly time-sensitive. Users must act within a narrow window before a game begins, making match recommendation a time-critical task to prevent missed engagement and revenue loss. Existing recommender systems, typically designed for static item catalogs, are ill-equipped to handle the hard temporal deadlines inherent in these live events. To address this, we designed and deployed a recommendation engine using the Deep Interest Network (DIN) architecture. We adapt the DIN architecture by injecting temporality at two levels: first, through real-time urgency features for each candidate match (e.g., time-to-round-lock), and second, via temporal positional encodings that represent the time-gap between each historical interaction and the current recommendation request, allowing the model to dynamically weigh the recency of past actions. This approach, combined with a listwise neuralNDCG loss function, produces highly relevant and urgency-aware rankings. To support this at industrial scale, we developed a multi-node, multi-GPU training architecture on Ray and PyTorch. Our system, validated on a massive industrial dataset with over 650k users and over 100B interactions, achieves a +9% lift in nDCG@1 over a heavily optimized LightGBM baseline with handcrafted features. The strong offline performance of this model establishes its viability as a core component for our planned on-device (edge) recommendation system, where online A/B testing will be conducted.

Yabo Ni, Cao Yuanpeng, Wenhang Zhou, Bangyang Hong, Zhongyi Zhang, Enlei Cai, Kangle Wu, Anxiang Zeng, Han Yu, Xiaoxiao Li

Large language models (LLMs) have shown great potential in enhancing search and recommender systems by providing rich semantic representations from unstructured texts. However, directly integrating LLM embeddings into industrial recommendation pipelines often results in subpar performance due to the semantic and distributional mismatch between pre-trained LLM features and domain-specific, feedback-driven representations. Existing approaches struggle to effectively align LLM embeddings with recommendation objectives, often facing challenges such as label misalignment or the potential loss of semantic diversity during fine-tuning. In this work, we present TreeBridge, a novel framework that introduces a structure-aware generative encoding tree to bridge the semantic gap between LLM embeddings and recommendation tasks. It preserves the external semantic richness of LLM embeddings, while learning label-informed structures that capture user preferences and interaction patterns. This enables the generation of task-adaptive representations without compromising embedding diversity. We further adopt an online-offline hybrid service paradigm to ensure low-latency real-world deployment. TreeBridge has been deployed on the Shopee e-commerce platform, one of the largest online shopping platforms in Southeast Asia serving hundreds of millions of users. Since its deployment in May 2025, it has helped the company achieve a commercially significant 1.55% relative improvement in gross merchandise volume (GMV). The deployment experience demonstrates the effectiveness, scalability, and significant commercial value of TreeBridge.

Mark Moussa, Amber V. Young, Brianna Isola, Vasuda Trehan, Michael D. Himes, Nicholas Wogan, Giada Arney

Future direct-imaging flagship missions, such as NASA's Habitable Worlds Observatory (HWO), face critical decisions in prioritizing observations due to extremely stringent time and resource constraints. In this paper, we introduce two advanced machine-learning architectures tailored for predicting biosignature species fluxes from exoplanetary reflected-light spectra: a Bayesian Convolutional Neural Network (BCNN) and our novel model architecture, the Spectral Query Adaptive Transformer (SQuAT). The BCNN robustly quantifies both epistemic and aleatoric uncertainties, offering reliable predictions under diverse observational conditions, whereas SQuAT employs query-driven attention mechanisms to enhance interpretability by explicitly associating spectral features with specific biosignature species. We demonstrate that both models achieve comparably high predictive accuracy on an augmented dataset spanning a wide range of exoplanetary conditions, while highlighting their distinct advantages in uncertainty quantification and spectral interpretability. These capabilities position our methods as promising tools for accelerating target triage, optimizing observation schedules, and maximizing scientific return for upcoming flagship missions such as HWO.

Gao Mo, Naveen Janaki Raman, Megan Chai, Cindy Peng, Shannon Pagdon, Nev Jones, Hong Shen, Margaret Swarbrick, Fei Fang

Behavioral health conditions, which include mental health and substance use disorders, are the leading disease burden in the United States. Peer-run behavioral health organizations (PROs) critically assist individuals facing these conditions by combining mental health services with assistance for needs such as income, employment, and housing. However, limited funds and staffing make it difficult for PROs to address all service user needs. To assist peer providers at PROs with their day-to-day tasks, we introduce PeerCoPilot, a large language model (LLM)-powered assistant that helps peer providers create wellness plans, construct step-by-step goals, and locate organizational resources to support these goals. PeerCoPilot ensures information reliability through a retrieval-augmented generation pipeline backed by a large database of over 1,300 vetted resources. We conducted human evaluations with 15 peer providers and 6 service users and found that over 90% of users supported using PeerCoPilot. Moreover, we demonstrate that PeerCoPilot provides more reliable and specific information than a baseline LLM. PeerCoPilot is now used by a group of 5-10 peer providers at CSPNJ, a large behavioral health organization serving over 10,000 service users, and we are actively expanding PeerCoPilot's use.

Pratik Mishra, Caner Gözübüyük, Seema Nagar, Prateeti Mohapatra, Raya Wittich, Arthur De Magalhaes

Manual creation of IT monitoring dashboard widgets is slow, error-prone, and a barrier for both novice and expert users. We present NOVAID, an interactive chatbot that leverages Large Language Models (LLMs) to generate IT monitoring widgets directly from natural language queries. Unlike general natural language–to-visualization tools, NOVAID addresses IT operations–specific challenges: specialized widget types like SLO charts, dynamic API-driven data retrieval, and complex contextual filters. The system combines a domain-aware semantic parser, fuzzy entity matching, and schema completion to produce standardized widget JSON specifications. An interactive clarification loop ensures accuracy in underspecified queries. On a curated dataset of 271 realistic queries, NOVAID achieves promising accuracy (up to 94.10% in metric extraction) across multiple LLMs. A user study with IT engineers yielded a System Usability Scale score of 74.2 for NOVAID, indicating good usability. By bridging natural language intent with operational dashboards, NOVAID demonstrates clear potential and a path for deployment in enterprise ITOps monitoring platforms.

Thomas Manzini, Priyankari Perali, Robin R. Murphy

This paper presents the first AI/ML system for automating building damage assessment in uncrewed aerial systems (sUAS) imagery to be deployed operationally during federally declared disasters (Hurricanes Debby and Helene). In response to major disasters, sUAS teams are dispatched to collect imagery of the affected areas to assess damage; however, at recent disasters, teams collectively delivered between 47GB and 369GB of imagery per day, representing more imagery than can reasonably be transmitted or interpreted by subject matter experts in the disaster scene, thus delaying response efforts. To alleviate this data avalanche encountered in practice, computer vision and machine learning techniques are necessary. While prior work has been deployed to automatically assess damage in satellite imagery, there is no current state of practice for sUAS-based damage assessment systems, as all known work has been confined to academic settings. This work establishes the state of practice via the development and deployment of models for building damage assessment with sUAS imagery. The model development involved training on the largest known dataset of post-disaster sUAS aerial imagery, containing 21,716 building damage labels, and the operational training of 91 disaster practitioners. The best performing model was deployed during the responses to Hurricanes Debby and Helene, where it assessed a combined 415 buildings in approximately 18 minutes. This work contributes documentation of the actual use of AI/ML for damage assessment during a disaster and lessons learned to the benefit of the AI/ML research and user communities.

Frederik Rajiv Manichand, Robin Deuber, Robert Jakob, Steve Swerling, Jamie Rosen, Elgar Fleisch, Patrick Langer

Estimating Body Mass Index (BMI) from camera images with machine learning models enables rapid weight assessment when traditional methods are unavailable or impractical, such as in telehealth or emergency scenarios. Existing computer vision approaches have been limited to datasets of up to 14,500 images. In this study, we present a deep learning-based BMI estimation method trained on our WayBED dataset, a large proprietary collection of 84,963 smartphone images from 25,353 individuals. We introduce an automatic filtering method that uses posture clustering and person detection to curate the dataset by removing low-quality images, such as those with atypical postures or incomplete views. This process retained 71,322 high-quality images suitable for training. We achieve a Mean Absolute Percentage Error (MAPE) of 7.9% on our hold-out test set (WayBED data) using full-body images, the lowest value in the published literature to the best of our knowledge. Further, we achieve a MAPE of 13% on the completely unseen (during training) VisualBodyToBMI dataset, comparable with state-of-the-art approaches trained on it, demonstrating robust generalization. Lastly, we fine-tune our model on VisualBodyToBMI and achieve a MAPE of 8.56%, the lowest reported value on this dataset so far. We deploy the full pipeline, including image filtering and BMI estimation, on Android devices using the CLAID framework. We release our complete code for model training, filtering, and the CLAID package for mobile deployment as open-source contributions.

Sel Ly, Rufan Yang, Ninad Dixit, Hung Dinh Nguyen

Lithium-ion (Li-ion) batteries are the major type of battery used in a variety of everyday applications, including electric vehicles (EVs), mobile devices, and energy storage systems. Predicting the Remaining Useful Life (RUL) of Li-ion batteries is crucial for ensuring their reliability, safety, and cost-effectiveness in battery-powered systems. The materials used for the battery cathodes and their designs play a significant role in determining the degradation rates and RUL, as they lead to distinct electrochemical reactions. Unfortunately, RUL prediction models often overlook the cathode materials and designs to simplify the model-building process, ignoring the effects of these electrochemical reactions. Other reasons are that specifications related to battery materials may not always be readily available, and a battery might consist of a mix of different materials. As a result, the predictive models that are developed often lack generalizability. To tackle these challenges, this paper proposes a novel material-based Mixture-of-Experts (MoE) approach for predicting the RUL of batteries, specifically addressing the complexities associated with heterogeneous battery chemistries. The MoE is integrated into a probabilistic framework, called Multiple Non-crossing Quantile Mixture-of-Experts for Probabilistic Prediction (RUL-QMoE), which accommodates battery operational conditions and enables uncertainty quantification. The RUL-QMoE model integrates specialized expert networks for five battery types: LFP, NCA, NMC, LCO, and NMC-LCO, within a gating mechanism that dynamically assigns relevance based on the battery's input features. Furthermore, by leveraging non-crossing quantile regression, the proposed RUL-QMoE produces coherent and interpretable predictive distributions of the battery's RUL, enabling robust uncertainty quantification in the battery's RUL prediction. Trained on seven real-world datasets, the proposed RUL-QMoE achieves strong predictive performance across all battery types, with MAE = 65 (cycles), MAPE = 9.59%, RMSE = 100 (cycles), and R2=96.84%. Compared to traditional models like XGBoost, Random Forest, CNN, and LSTM, the proposed RUL-QMoE model consistently delivers lower RMSE and superior probabilistic insights, including survival probabilities and prediction intervals. The model has been integrated into our Battery AI platform in collaboration with Toyota Motor Engineering & Manufacturing North America, Inc., as part of a broader Battery Foundation Model initiative. This RUL-QMoE model will serve future Toyota EVs' users and battery system designers.

Wenjie Lin, Jin Wei-Kocsis

While large language models (LLMs) have shown promise in healthcare, their application for rare medical conditions is still hindered by scarce and unreliable datasets for fine-tuning. Hyperhidrosis, a disorder causing excessive sweating beyond physiological needs, is one such rare disorder, affecting 2-3% of the population and significantly impacting both physical comfort and psychosocial well-being. To date, no work has tailored LLMs to advance the diagnosis or care of hyperhidrosis. To address this gap, we present LLM4Sweat, an open-source and domain-specific LLM framework for trustworthy and empathetic hyperhidrosis support. The system follows a three-stage pipeline. In the data augmentation stage, a frontier LLM generates medically plausible synthetic vignettes from curated open-source data to create a diverse and balanced question–answer dataset. In the fine-tuning stage, an open-source foundation model is fine-tuned on the dataset to provide diagnosis, personalized treatment recommendations, and empathetic psychological support. In the inference and expert evaluation stage, clinical and psychological specialists assess accuracy, appropriateness, and empathy, with validated responses iteratively enriching the dataset. Experiments show that LLM4Sweat outperforms baselines and delivers the first open-source LLM framework for hyperhidrosis, offering a generalizable approach for other rare diseases with similar data and trustworthiness challenges.

Youngkyu Lee, Jinho Lee, Youngdae Jo, Jeongwoo Park

Timely detection of retinal diseases is crucial for preventing vision loss; yet the limited availability of ophthalmologists and disparities in access to diagnostic services continue to hinder widespread screening, particularly in primary care settings. We present REMEDIS, a Software-as-a-Service (SaaS)-based clinical AI framework for the automated diagnosis of major retinal diseases, including age-related macular degeneration (AMD), diabetic retinopathy (DR), epiretinal membrane (ERM), and glaucoma, using fundus images. The system analyzes high-resolution fundus photographs in a secure cloud environment via a Swin-Large-based multi-disease classification network, producing disease-specific probability scores. To ensure clinically meaningful decision making, Youden’s Index is applied to determine optimized sensitivity-specificity thresholds for each condition. An explainability module based on Grad-CAM generates lesion localization contour visualizations, providing interpretable evidence that assists ophthalmologists in case review and facilitates integration into electronic medical records (EMR). The framework was evaluated in an IRB-approved multicenter prospective clinical trial conducted under real-world conditions, achieving an average AUC exceeding 0.94 across the four target diseases and demonstrating strong concordance with expert diagnoses. To our knowledge, this represents one of the first SaaS-based AI diagnostic frameworks for retinal diseases validated through prospective clinical studies, highlighting its potential as an emerging clinical application of AI.

Valerio La Gatta, Marco Postiglione, Jeremy Gilbert, Daniel W. Linna Jr., Morgan Manella Greenfield, Aaron Shaw, VS Subrahmanian

Numerous social movements (SMs) around the world help support the UN's Sustainable Development Goals (SDGs). Understanding how key events shape SMs is key to the achievement of the SDGs. We have developed SMART (Social Media Analysis & Reasoning Tool) to track social movements related to the SDGs. SMART was designed by a multidisciplinary team of AI researchers, journalists, communications scholars and legal experts. This paper describes SMART's transformer-based multivariate time series Discourse Evolution Engine for Predictions about Social Movements (DEEP) to predict the volume of future articles/posts and the emotions expressed. DEEP outputs probabilistic forecasts with uncertainty estimates, providing critical support for editorial planning and strategic decision-making. We evaluate DEEP with a case study of the #MeToo movement by creating a novel longitudinal dataset (433K Reddit posts and 121K news articles) from September 2024 to June 2025, which is publicly available for research purposes.

Tarun Kumar, Aalap Tripathy, Gayathri Saranathan, Martin Foltin, Suparna Bhattacharya, Scott Hinchley, Donald M Bahls, David Brookshire, Larry Kaplan, Robert W. Wisniewski

The proliferation of Model Context Protocol (MCP) servers in enterprise infrastructure management has revolutionized AI-driven automation while introducing critical multi-layered security vulnerabilities that traditional cybersecurity frameworks cannot adequately address. This paper presents a comprehensive intelligent guardrail system that addresses the unique security challenges of MCP-driven infrastructure management through a novel four-layer defense architecture. Our solution employs a dedicated guardian LLM that interprets natural language policies and applies contextual reasoning to complex infrastructure scenarios, providing dynamic policy enforcement that adapts to user roles, operational timing, and system context. Unlike existing rule-based security systems, our approach implements guardrails at four distinct control points: input message filtering, tool selection validation, execution-time verification, and post-action auditing. The system addresses critical gaps in existing security solutions by providing infrastructure-specific threat modeling, real-time policy adaptation, and comprehensive audit trails with explainable decision-making through confidence scores and detailed reasoning. Our evaluation demonstrates the system's effectiveness in preventing command injection, privilege escalation, and tool poisoning attacks across various enterprise infrastructure scenarios while maintaining operational agility essential for modern data center management.

Aman Kumar, Ekant Muljibhai Amin, Xian Yeow Lee, Lasitha Vidyaratne, Ahmed K. Farahat, Dipanjan D. Ghosh, Yuta Koreeda, Chetan Gupta

Large Language Models (LLMs) have shown remarkable success in supporting a wide range of knowledge-intensive tasks. In specialized domains, there is growing interest in leveraging LLMs to assist subject matter experts with domain-specific challenges. However, deploying LLMs as SaaS solutions raises data privacy concerns, while many open-source models demand significant computational resources for effective domain adaptation and deployment. A promising alternative is to develop smaller, domain-specialized LLMs, though this approach is often constrained by the lack of high-quality domain-specific training data. In this work, we address these limitations by presenting a cost-efficient and scalable training pipeline that combines guided synthetic data generation from a small seed corpus with bottom-up domain data curation. Our pipeline integrates Domain-Adaptive Pretraining (DAPT), Domain-specific Supervised Fine-tuning (DSFT), and Direct Preference Optimization (DPO) to train effective small-scale models for specialized use cases. We demonstrate this approach through DiagnosticSLM, a 3B-parameter domain-specific model tailored for fault diagnosis, root cause analysis, and repair recommendation in industrial settings. To evaluate model performance, we introduce four domain-specific benchmarks: multiple-choice questions (DiagnosticMCQ), question answering (DiagnosticQA), sentence completion (DiagnosticComp), and summarization (DiagnosticSum). DiagnosticSLM achieves up to 25% accuracy improvement over open-source models of comparable or larger size (2B-9B) on the MCQ task, while also outperforming or matching them in other tasks, demonstrating effective domain-specific reasoning and generalization capabilities.

Lingkai Kong, Haichuan Wang, Charles A. Emogor, Vincent Böersch-Supan, Lily Xu, Milind Tambe

Poaching poses significant threats to biodiversity. A valuable step in reducing poaching is to forecast poacher behavior, which can inform patrol deployment and other conservation interventions. Existing poaching prediction methods based on linear models or decision trees lack the expressivity to capture complex, nonlinear spatiotemporal patterns. Recent advances in generative modeling, particularly flow matching, offer a more flexible alternative. However, training such models on real-world poaching data faces two central obstacles: imperfect detection of poaching events and limited data. To address imperfect detection, we integrate flow matching with an occupancy-based detection model and train the flow in latent space to infer the underlying occupancy state. To mitigate data scarcity, we adopt a composite flow initialized from a linear-model prediction rather than random noise which is the standard in diffusion models, injecting prior knowledge and improving generalization. Evaluations on datasets from two national parks in Uganda show consistent gains in predictive accuracy.

Jisu Kim, Mahimul Islam, Ashok Goel

The ability to correct mistakes and adapt to users' changing needs is critical for AI agents to remain robust and trustworthy. LLM-based agents are inherently prone to errors like hallucinations and misinterpretations. We observed this challenge in SAMI, an AI social agent deployed in Georgia Tech's OMSCS program for ten semesters (11,000+ users). Users frequently requested the agent to revise its knowledge base, both to correct LLM-induced errors and to update their information. To support such revisions, we introduce a two-level metacognitive self-adaptation architecture that integrates knowledge-based AI (KBAI) with LLMs. The architecture comprises a cognitive layer that performs the agent's core tasks, and a metacognitive layer that introspects on the cognitive layer's process using a Task–Method–Knowledge (TMK) model of the agent. The metacognitive layer identifies the task that needs revision, updates the knowledge base, and communicates the revision process to the user.

Qiming Guo, Bishal Khatri, Wenbo Sun, Jinwen Tang, Hua Zhang, Wenlu Wang

Underground pipeline leaks and infiltrations pose significant threats to water security and environmental safety. Traditional manual inspection methods provide limited coverage and delayed response, often missing critical anomalies. This paper proposes AquaSentinel, a novel physics-informed AI system for real-time anomaly detection in urban underground water pipeline networks. We introduce four key innovations: (1) strategic sparse sensor deployment at high-centrality nodes combined with physics-based state augmentation to achieve network-wide observability from minimal infrastructure; (2) the RTCA (Real-Time Cumulative Anomaly) detection algorithm, which employs dual-threshold monitoring with adaptive statistics to distinguish transient fluctuations from genuine anomalies; (3) a Mixture of Experts (MoE) ensemble of spatiotemporal graph neural networks that provides robust predictions by dynamically weighting model contributions; (4) causal flow-based leak localization that traces anomalies upstream to identify source nodes and affected pipe segments. Our system strategically deploys sensors at critical network junctions and leverages physics-based modeling to propagate measurements to unmonitored nodes, creating virtual sensors that enhance data availability across the entire network. Experimental evaluation using 110 leak scenarios demonstrates that AquaSentinel achieves 100% detection accuracy. This work advances pipeline monitoring by demonstrating that physics-informed sparse sensing can match the performance of dense deployments at a fraction of the cost, providing a practical solution for aging urban infrastructure.

Haeji Go, Jae-Hun Lee, Shinyeong Noh, Kayoung Kim, Kyuseong Lim, Jee Eun Song, Mingyu Lee, Joowan Sung, Soonbeom Kwon, Myoungbok Shin 等

Designing vehicle exteriors requires repeated refinement of key proportions and viewpoints, a process traditionally reliant on manual sketching, which is often time-consuming and inefficient in early concept stages. To accelerate the design process, we are exploring the potential of utilizing AI for ideation in these early stages. However, it remains a challenging task to control proportions and maintain a fixed perspective when generating images using AI. To address these limitations, we present RatioMorph, a controllable image generation system that enables manipulation of vehicle proportions and viewpoints when generating images by AI. RatioMorph comprises two core modules. Car2BoxNet is a depth estimation model that transforms real photographs into structured box-style depth maps that capture the geometric layout of the vehicle. Box2CarNet is a diffusion-based image generator fine-tuned to produce vehicle designs that adhere to the provided geometric conditions. Both Car2BoxNet and Box2CarNet are trained on a synthetic dataset curated through automated filtering based on geometric alignment and visual quality. Evaluated within a production-adjacent automotive design workflow, RatioMorph significantly reduced early-stage design iteration time and enabled exploratory workflows that were difficult with previous AI workflows. This work introduces a domain-specific, controllable diffusion-based generation system tailored for automotive design, enabling manipulation of vehicle viewpoint and proportion. It demonstrates strong potential to accelerate early-stage workflows and outlines a path toward industrial deployment, with phased integration into production environments currently underway.

Taylor Dinkins, Weng-Keen Wong, Basavaraj Amogi, Paola Pesantez-Cabrera, Jaitun Patel, Lav Khot, Alan Fern

Near surface temperature inversions are periods in which a low layer of warm air is trapped between cooler air higher up in the atmosphere and dense cooler air below it near the surface level. By causing cooler air to pool near the surface level, inversions can have detrimental effects for crop growers, including frost, increased moisture, and pesticide drift. As a result, predicting the occurrence and magnitude of these inversions yields substantial benefits for growers. We introduce a Long Short-Term Memory (LSTM) model for temperature inversion forecasting that is able to effectively predict localized, near surface temperature inversions in advance such that growers can take actions to mitigate the detrimental effects. We show a substantial performance gain over a deployed temperature inversion forecasting system, and include a series of ablations that show the benefit of using publicly available terrain-specific feature information when modeling inversions at this scale.