Restoring Real-World Images with an Internal Detail Enhancement Diffusion Model

Restoring real-world degraded images, such as old photographs or low-resolution images, presents a significant challenge due to the complex, mixed degradations they exhibit, such as scratches, color fading, and noise. Recent data-driven approaches have struggled with two main challenges: achieving high-fidelity restoration and providing object-level control over colorization. While diffusion models have shown promise in generating high-quality images with specific controls, they often fail to fully preserve image details during restoration. In this work, we propose an internal detail-preserving diffusion model for high-fidelity restoration of real-world degraded images. Our method utilizes a pre-trained Stable Diffusion model as a generative prior, eliminating the need to train a model from scratch. Central to our approach is the Internal Image Detail Enhancement (IIDE) technique, which directs the diffusion model to preserve essential structural and textural information while mitigating degradation effects. The process starts by mapping the input image into a latent space, where we inject the diffusion denoising process with degradation operations that simulate the effects of various degradation factors. Extensive experiments demonstrate that our method significantly outperforms state-of-the-art models in both qualitative assessments and perceptual quantitative evaluations. Additionally, our approach supports text-guided restoration, enabling object-level colorization control that mimics the expertise of professional photo editing.

RemoteSAM: Towards Segment Anything for Earth Observation

We aim to develop a robust yet flexible visual foundation model for Earth observation. It should possess strong capabilities in recognizing and localizing diverse visual targets while providing compatibility with various input-output interfaces required across different task scenarios. Current systems cannot meet these requirements, as they typically utilize task-specific architecture trained on narrow data domains with limited semantic coverage. Our study addresses these limitations from two aspects: data and modeling. We first introduce an automatic data engine that enjoys significantly better scalability compared to previous human annotation or rule-based approaches. It has enabled us to create the largest dataset of its kind to date, comprising 270K image-text-mask triplets covering an unprecedented range of diverse semantic categories and attribute specifications. Based on this data foundation, we further propose a task unification paradigm that centers around referring expression segmentation. It effectively handles a wide range of vision-centric perception tasks, including classification, detection, segmentation, grounding, etc, using a single model without any task-specific heads. Combining these innovations on data and modeling, we present RemoteSAM, a foundation model that establishes new SoTA on several earth observation perception benchmarks, outperforming other foundation models such as Falcon, GeoChat, and LHRS-Bot with significantly higher efficiency. Models and data are publicly available at https://github.com/1e12Leon/RemoteSAM.

HARDMath2: A Benchmark for Applied Mathematics Built by Students as Part of a Graduate Class

Large language models (LLMs) have shown remarkable progress in mathematical problem-solving, but evaluation has largely focused on problems that have exact analytical solutions or involve formal proofs, often overlooking approximation-based problems ubiquitous in applied science and engineering. To fill this gap, we build on prior work and present HARDMath2, a dataset of 211 original problems covering the core topics in an introductory graduate applied math class, including boundary-layer analysis, WKB methods, asymptotic solutions of nonlinear partial differential equations, and the asymptotics of oscillatory integrals. This dataset was designed and verified by the students and instructors of a core graduate applied mathematics course at Harvard. We build the dataset through a novel collaborative environment that challenges students to write and refine difficult problems consistent with the class syllabus, peer-validate solutions, test different models, and automatically check LLM-generated solutions against their own answers and numerical ground truths. Evaluation results show that leading frontier models still struggle with many of the problems in the dataset, highlighting a gap in the mathematical reasoning skills of current LLMs. Importantly, students identified strategies to create increasingly difficult problems by interacting with the models and exploiting common failure modes. This back-and-forth with the models not only resulted in a richer and more challenging benchmark but also led to qualitative improvements in the students' understanding of the course material, which is increasingly important as we enter an age where state-of-the-art language models can solve many challenging problems across a wide domain of fields.

MedGUIDE: Benchmarking Clinical Decision-Making in Large Language Models

Clinical guidelines, typically structured as decision trees, are central to evidence-based medical practice and critical for ensuring safe and accurate diagnostic decision-making. However, it remains unclear whether Large Language Models (LLMs) can reliably follow such structured protocols. In this work, we introduce MedGUIDE, a new benchmark for evaluating LLMs on their ability to make guideline-consistent clinical decisions. MedGUIDE is constructed from 55 curated NCCN decision trees across 17 cancer types and uses clinical scenarios generated by LLMs to create a large pool of multiple-choice diagnostic questions. We apply a two-stage quality selection process, combining expert-labeled reward models and LLM-as-a-judge ensembles across ten clinical and linguistic criteria, to select 7,747 high-quality samples. We evaluate 25 LLMs spanning general-purpose, open-source, and medically specialized models, and find that even domain-specific LLMs often underperform on tasks requiring structured guideline adherence. We also test whether performance can be improved via in-context guideline inclusion or continued pretraining. Our findings underscore the importance of MedGUIDE in assessing whether LLMs can operate safely within the procedural frameworks expected in real-world clinical settings.

Dome-DETR: DETR with Density-Oriented Feature-Query Manipulation for Efficient Tiny Object Detection

Tiny object detection plays a vital role in drone surveillance, remote sensing, and autonomous systems, enabling the identification of small targets across vast landscapes. However, existing methods suffer from inefficient feature leverage and high computational costs due to redundant feature processing and rigid query allocation. To address these challenges, we propose Dome-DETR, a novel framework with Density-Oriented Feature-Query Manipulation for Efficient Tiny Object Detection. To reduce feature redundancies, we introduce a lightweight Density-Focal Extractor (DeFE) to produce clustered compact foreground masks. Leveraging these masks, we incorporate Masked Window Attention Sparsification (MWAS) to focus computational resources on the most informative regions via sparse attention. Besides, we propose Progressive Adaptive Query Initialization (PAQI), which adaptively modulates query density across spatial areas for better query allocation. Extensive experiments demonstrate that Dome-DETR achieves state-of-the-art performance (+3.3 AP on AI-TOD-V2 and +2.5 AP on VisDrone) while maintaining low computational complexity and a compact model size. Code will be released upon acceptance.

SelaFD:Seamless Adaptation of Vision Transformer Fine-tuning for Radar-based Human Activity

Human Activity Recognition (HAR) such as fall detection has become increasingly critical due to the aging population, necessitating effective monitoring systems to prevent serious injuries and fatalities associated with falls. This study focuses on fine-tuning the Vision Transformer (ViT) model specifically for HAR using radar-based Time-Doppler signatures. Unlike traditional image datasets, these signals present unique challenges due to their non-visual nature and the high degree of similarity among various activities. Directly fine-tuning the ViT with all parameters proves suboptimal for this application. To address this challenge, we propose a novel approach that employs Low-Rank Adaptation (LoRA) fine-tuning in the weight space to facilitate knowledge transfer from pre-trained ViT models. Additionally, to extract fine-grained features, we enhance feature representation through the integration of a serial-parallel adapter in the feature space. Our innovative joint fine-tuning method, tailored for radar-based Time-Doppler signatures, significantly improves HAR accuracy, surpassing existing state-of-the-art methodologies in this domain. Our code is released at https://github.com/wangyijunlyy/SelaFD.

B-STaR: Monitoring and Balancing Exploration and Exploitation in Self-Taught Reasoners

In the absence of extensive human-annotated data for complex reasoning tasks, self-improvement -- where models are trained on their own outputs -- has emerged as a primary method for enhancing performance. However, the critical factors underlying the mechanism of these iterative self-improving methods remain poorly understood, such as under what conditions self-improvement is effective, and what are the bottlenecks in the current iterations. In this work, we identify and propose methods to monitor two pivotal factors in this iterative process: (1) the model's ability to generate sufficiently diverse responses (exploration); and (2) the effectiveness of external rewards in distinguishing high-quality candidates from lower-quality ones (exploitation). Using mathematical reasoning as a case study, we begin with a quantitative analysis to track the dynamics of exploration and exploitation, discovering that a model's exploratory capabilities rapidly deteriorate over iterations, and the effectiveness of exploiting external rewards diminishes as well. Motivated by these findings, we introduce B-STaR, a Self-Taught Reasoning framework that autonomously adjusts configurations across iterations to Balance exploration and exploitation, thereby optimizing the self-improving effectiveness based on the current policy model and available rewards. Our experiments on mathematical reasoning, coding, and commonsense reasoning demonstrate that B-STaR not only enhances the model's exploratory capabilities throughout training but also achieves a more effective balance between exploration and exploitation, leading to superior performance.

Learning to Denoise Unreliable Interactions for Link Prediction on Biomedical Knowledge Graph

Link prediction in biomedical knowledge graphs (KGs) aims at predicting unknown interactions between entities, including drug-target interaction (DTI) and drug-drug interaction (DDI), which is critical for drug discovery and therapeutics. Previous methods prefer to utilize the rich semantic relations and topological structure of the KG to predict missing links, yielding promising outcomes. However, all these works only focus on improving the predictive performance without considering the inevitable noise and unreliable interactions existing in the KGs, which limits the development of KG-based computational methods. To address these limitations, we propose a Denoised Link Prediction framework, called DenoisedLP. DenoisedLP obtains reliable interactions based on the local subgraph by denoising noisy links in a learnable way, providing a universal module for mining underlying task-relevant relations. To collaborate with the smoothed semantic information, DenoisedLP introduces the semantic subgraph by blurring conflict relations around the predicted link. By maximizing the mutual information between the reliable structure and smoothed semantic relations, DenoisedLP emphasizes the informative interactions for predicting relation-specific links. Experimental results on real-world datasets demonstrate that DenoisedLP outperforms state-of-the-art methods on DTI and DDI prediction tasks, and verify the effectiveness and robustness of denoising unreliable interactions on the contaminated KGs.

Visual tracking brain computer interface

Brain-computer interfaces (BCIs) offer a way to interact with computers without relying on physical movements. Non-invasive electroencephalography (EEG)-based visual BCIs, known for efficient speed and calibration ease, face limitations in continuous tasks due to discrete stimulus design and decoding methods. To achieve continuous control, we implemented a novel spatial encoding stimulus paradigm and devised a corresponding projection method to enable continuous modulation of decoded velocity. Subsequently, we conducted experiments involving 17 participants and achieved Fitt's ITR of 0.55 bps for the fixed tracking task and 0.37 bps for the random tracking task. The proposed BCI with a high Fitt's ITR was then integrated into two applications, including painting and gaming. In conclusion, this study proposed a visual BCI-based control method to go beyond discrete commands, allowing natural continuous control based on neural activity.

High-performance cVEP-BCI under minimal calibration

The ultimate goal of brain-computer interfaces (BCIs) based on visual modulation paradigms is to achieve high-speed performance without the burden of extensive calibration. Code-modulated visual evoked potential-based BCIs (cVEP-BCIs) modulated by broadband white noise (WN) offer various advantages, including increased communication speed, expanded encoding target capabilities, and enhanced coding flexibility. However, the complexity of the spatial-temporal patterns under broadband stimuli necessitates extensive calibration for effective target identification in cVEP-BCIs. Consequently, the information transfer rate (ITR) of cVEP-BCI under limited calibration usually stays around 100 bits per minute (bpm), significantly lagging behind state-of-the-art steady-state visual evoked potential-based BCIs (SSVEP-BCIs), which achieve rates above 200 bpm. To enhance the performance of cVEP-BCIs with minimal calibration, we devised an efficient calibration stage involving a brief single-target flickering, lasting less than a minute, to extract generalizable spatial-temporal patterns. Leveraging the calibration data, we developed two complementary methods to construct cVEP temporal patterns: the linear modeling method based on the stimulus sequence and the transfer learning techniques using cross-subject data. As a result, we achieved the highest ITR of 250 bpm under a minute of calibration, which has been shown to be comparable to the state-of-the-art SSVEP paradigms. In summary, our work significantly improved the cVEP performance under few-shot learning, which is expected to expand the practicality and usability of cVEP-BCIs.