Searcharxiv⌕ Search

arXiv subjects

Fan Zhang

Publications and source records attributed to Fan Zhang.

At least 235 records · Page 13Linked to original sources
arXiv2026

Few-Shot Continual Learning for 3D Brain MRI with Frozen Foundation Models

Foundation models pretrained on large-scale 3D medical imaging data face challenges when adapted to multiple downstream tasks under continual learning with limited labeled data. We address few-shot continual learning for 3D brain MRI by combining a frozen pretrained backbone with task-specific Low-Rank Adaptation (LoRA) modules. Tasks arrive sequentially -- tumor segmentation (BraTS) and brain age estimation (IXI) -- with no replay of previous task data. Each task receives a dedicated LoRA adapter; only the adapter and task-specific head are trained while the backbone remains frozen, thereby eliminating catastrophic forgetting by design (BWT=0). In continual learning, sequential full fine-tuning suffers severe forgetting (T1 Dice drops from 0.80 to 0.16 after T2), while sequential linear probing achieves strong T1 (Dice 0.79) but fails on T2 (MAE 1.45). Our LoRA approach achieves the best balanced performance across both tasks: T1 Dice 0.62$\pm$0.07, T2 MAE 0.16$\pm$0.05, with zero forgetting and $<$0.1\% trainable parameters per task, though with noted systematic age underestimation in T2 (Wilcoxon $p<0.001$). Frozen foundation models with task-specific LoRA adapters thus offer a practical solution when both tasks must be maintained under few-shot continual learning.

eess.IV↗
arXiv2026

Trajectory-aware Shifted State Space Models for Online Video Super-Resolution

Online video super-resolution (VSR) is an important technique for many real-world video processing applications, which aims to restore the current high-resolution video frame based on temporally previous frames. Most of the existing online VSR methods solely employ one neighboring previous frame to achieve temporal alignment, which limits long-range temporal modeling of videos. Recently, state space models (SSMs) have been proposed with linear computational complexity and a global receptive field, which significantly improve computational efficiency and performance. In this context, this paper presents a novel online VSR method based on Trajectory-aware Shifted SSMs (TS-Mamba), leveraging both long-term trajectory modeling and low-complexity Mamba to achieve efficient spatio-temporal information aggregation. Specifically, TS-Mamba first constructs the trajectories within a video to select the most similar tokens from the previous frames. Then, a Trajectory-aware Shifted Mamba Aggregation (TSMA) module consisting of proposed shifted SSMs blocks is employed to aggregate the selected tokens. The shifted SSMs blocks are designed based on Hilbert scannings and corresponding shift operations to compensate for scanning losses and strengthen the spatial continuity of Mamba. Additionally, we propose a trajectory-aware loss function to supervise the trajectory generation, ensuring the accuracy of token selection when training our model. Extensive experiments on three widely used VSR test datasets demonstrate that compared with six online VSR benchmark models, our TS-Mamba achieves state-of-the-art performance in most cases and over 22.7% complexity reduction (in MACs).

cs.CV↗
arXiv2026

Making Slow Thinking Faster: Compressing LLM Chain-of-Thought via Step Entropy

Large Language Models (LLMs) using Chain-of-Thought (CoT) prompting excel at complex reasoning but generate verbose thought processes with considerable redundancy, leading to increased inference costs and reduced efficiency. We introduce a novel CoT compression framework based on step entropy, a metric that quantifies \emph{the informational contribution of individual reasoning steps} to identify redundancy. Through theoretical analysis and extensive empirical validation on mathematical reasoning benchmarks, we demonstrate that steps with low entropy are indeed highly redundant. Our experiments reveal that an astonishing 80\% of low-entropy intermediate steps can be pruned with minor degradation in the final answer accuracy across DeepSeek-R1-7B, 14B and Qwen3-8B. This finding sharply contrasts with random or high-entropy pruning, which severely impairs reasoning performance. Building on this, we propose a novel two-stage training strategy combining Supervised Fine-Tuning (SFT) and Group Relative Policy Optimization (GRPO) reinforcement learning. This approach enables LLMs to autonomously learn to generate compressed COTs during inference by strategically incorporating [SKIP] tokens. Our method significantly improves LLM inference efficiency while preserving accuracy, paving the way for more scalable LLM deployments and a better understanding of their internal reasoning. The code and data are released in https://github.com/staymylove/COT_Compresstion_via_Step_entropy.

cs.AI↗
arXiv2026

Segment-Based Two-Loop Adaptive Iterative Learning Control for Spacecraft Position and Attitude Tracking

Proximity operations of rigid bodies, such as spacecraft rendezvous and docking, require precise tracking of both position and attitude over finite time intervals. These operations are often repeated under uncertain conditions, with unknown but repeatable parameters and disturbances. Adaptive iterative learning control (ILC) is well suited to such tasks, as it can track desired trajectories while learning unknown, iteration-invariant signals or parameters. However, conventional adaptive ILC faces two challenges: (i) the coupling between rotational and translational dynamics complicates the design of the two coordinated learning loops for position and attitude, and (ii) standard adaptive ILC designs cannot guarantee bounded control inputs. To address these issues, we propose a dual-number-based, segment-based two-loop adaptive ILC framework for simultaneous high-precision position and attitude tracking. The framework employs two learning loops that interact through a dual-number representation of tracking errors, combining position and attitude errors into a single mathematical object for unified control design. A segment-based dynamic projection mechanism ensures that both parameter estimates and control inputs remain bounded without prior knowledge of uncertainties. Mathematical analysis and numerical simulations demonstrate that the proposed framework significantly enhances tracking performance under unknown but repeatable uncertainties and strong rotational-translational coupling.

eess.SY↗
arXiv2026

S2R-HDR: A Large-Scale Rendered Dataset for HDR Fusion

The generalization of learning-based high dynamic range (HDR) fusion is often limited by the availability of training data, as collecting large-scale HDR images from dynamic scenes is both costly and technically challenging. To address these challenges, we propose S2R-HDR, the first large-scale high-quality synthetic dataset for HDR fusion, with 24,000 HDR samples. Using Unreal Engine 5, we design a diverse set of realistic HDR scenes that encompass various dynamic elements, motion types, high dynamic range scenes, and lighting. Additionally, we develop an efficient rendering pipeline to generate realistic HDR images. To further mitigate the domain gap between synthetic and real-world data, we introduce S2R-Adapter, a domain adaptation designed to bridge this gap and enhance the generalization ability of models. Experimental results on real-world datasets demonstrate that our approach achieves state-of-the-art HDR fusion performance. Dataset and code are available at https://openimaginglab.github.io/S2R-HDR.

cs.CV↗
arXiv2026

Quantum Temporal Convolutional Neural Networks for Cross-Sectional Equity Return Prediction: A Comparative Benchmark Study

Quantum machine learning offers a promising pathway for enhancing stock market prediction, particularly under complex, noisy, and highly dynamic financial environments. However, many classical forecasting models struggle with noisy input, regime shifts, and limited generalization capacity. To address these challenges, we propose a Quantum Temporal Convolutional Neural Network (QTCNN) that combines a classical temporal encoder with parameter-efficient quantum convolution circuits for cross-sectional equity return prediction. The temporal encoder extracts multi-scale patterns from sequential technical indicators, while the quantum processing leverages superposition and entanglement to enhance feature representation and suppress overfitting. We conduct a comprehensive benchmarking study on the JPX Tokyo Stock Exchange dataset and evaluate predictions through long-short portfolio construction using out-of-sample Sharpe ratio as the primary performance metric. QTCNN achieves a Sharpe ratio of 0.538, outperforming the best classical baseline by approximately 72\%. These results highlight the practical potential of quantum-enhanced forecasting model, QTCNN, for robust decision-making in quantitative finance.

cs.LG↗
arXiv2026

SAM3-LiteText: An Anatomical Study of the SAM3 Text Encoder for Efficient Vision-Language Segmentation

Vision-language segmentation models such as SAM3 enable flexible, prompt-driven visual grounding, but inherit large, general-purpose text encoders originally designed for open-ended language understanding. In practice, segmentation prompts are short, structured, and semantically constrained, leading to substantial over-provisioning in text encoder capacity and persistent computational and memory overhead. In this paper, we perform a large-scale anatomical analysis of text prompting in vision-language segmentation, covering 404,796 real prompts across multiple benchmarks. Our analysis reveals severe redundancy: most context windows are underutilized, vocabulary usage is highly sparse, and text embeddings lie on low-dimensional manifold despite high-dimensional representations. Motivated by these findings, we propose SAM3-LiteText, a lightweight text encoding framework that replaces the original SAM3 text encoder with a compact MobileCLIP student that is optimized by knowledge distillation. Extensive experiments on image and video segmentation benchmarks show that SAM3-LiteText reduces text encoder parameters by up to 88%, substantially reducing static memory footprint, while maintaining segmentation performance comparable to the original model. Code: https://github.com/SimonZeng7108/efficientsam3/tree/sam3_litetext.

cs.AI↗
arXiv2026

MME-Emotion: A Holistic Evaluation Benchmark for Emotional Intelligence in Multimodal Large Language Models

Recent advances in multimodal large language models (MLLMs) have catalyzed transformative progress in affective computing, enabling models to exhibit emergent emotional intelligence. Despite substantial methodological progress, current emotional benchmarks remain limited, as it is still unknown: (a) the generalization abilities of MLLMs across distinct scenarios, and (b) their reasoning capabilities to identify the triggering factors behind emotional states. To bridge these gaps, we present \textbf{MME-Emotion}, a systematic benchmark that assesses both emotional understanding and reasoning capabilities of MLLMs, enjoying \textit{scalable capacity}, \textit{diverse settings}, and \textit{unified protocols}. As the largest emotional intelligence benchmark for MLLMs, MME-Emotion contains over 6,000 curated video clips with task-specific questioning-answering (QA) pairs, spanning broad scenarios to formulate eight emotional tasks. It further incorporates a holistic evaluation suite with hybrid metrics for emotion recognition and reasoning, analyzed through a multi-agent system framework. Through a rigorous evaluation of 20 advanced MLLMs, we uncover both their strengths and limitations, yielding several key insights: \ding{182} Current MLLMs exhibit unsatisfactory emotional intelligence, with the best-performing model achieving only $39.3\%$ recognition score and $56.0\%$ Chain-of-Thought (CoT) score on our benchmark. \ding{183} Generalist models (\emph{e.g.}, Gemini-2.5-Pro) derive emotional intelligence from generalized multimodal understanding capabilities, while specialist models (\emph{e.g.}, R1-Omni) can achieve comparable performance through domain-specific post-training adaptation. By introducing MME-Emotion, we hope that it can serve as a foundation for advancing MLLMs' emotional intelligence in the future.

cs.CV↗
arXiv2026

The Effect of Design Thinking on Creative & Innovation Processes: An Empirical Study Across Different Design Experience Levels

This study employs linear regression and structural equation modeling to explore how Thinking Skills, Design Thinking, Creative Self-Efficacy (CSE), and Collective Creative Efficacy (CCE) drive Design Creativity & Innovation, and analyzes the structural stability of the model across different levels of experience. Path analysis results indicate that the four Design Thinking Skills, Problem-driven Design (beta = 0.198, p < 0.01), Information-driven Design (beta = 0.241, p < 0.001), Solution-driven Design (beta = 0.227, p < 0.001), and Knowledge-driven Design (beta = 0.263, p < 0.001) all significantly and positively influence Design Thinking. Furthermore, Design Thinking has a significant positive predictive effect on Design Creativity & Innovation (beta = 0.286, p < 0.001). Mediation analysis confirms three significant mediation paths: the CSE mediation path (beta = 0.128, p < 0.001), the CCE mediation path (beta = 0.073, p < 0.01), and the "CSE to CCE" chain mediation path (beta = 0.025, p < 0.01). Multi-group comparison results reveal significant differences between the student and professional groups under the full equivalence model. After relaxing specific constraints, there were no significant differences between the nested models of the baseline model, partial measurement invariance, structural weight invariance, and structural covariance invariance. These findings elucidate the multi-dimensional pathways of Design Creativity & Innovation, providing a robust empirical basis for optimizing differentiated pedagogical models and professional practice guidelines.

cs.HC↗
arXiv2026

The CLEF-2026 FinMMEval Lab: Multilingual and Multimodal Evaluation of Financial AI Systems

We present the setup and the tasks of the FinMMEval Lab at CLEF 2026, which introduces the first multilingual and multimodal evaluation framework for financial Large Language Models (LLMs). While recent advances in financial natural language processing have enabled automated analysis of market reports, regulatory documents, and investor communications, existing benchmarks remain largely monolingual, text-only, and limited to narrow subtasks. FinMMEval 2026 addresses this gap by offering three interconnected tasks that span financial understanding, reasoning, and decision-making: Financial Exam Question Answering, Multilingual Financial Question Answering (PolyFiQA), and Financial Decision Making. Together, these tasks provide a comprehensive evaluation suite that measures models' ability to reason, generalize, and act across diverse languages and modalities. The lab aims to promote the development of robust, transparent, and globally inclusive financial AI systems, with datasets and evaluation resources publicly released to support reproducible research.

cs.CL↗
arXiv2026

GECAM discovery of the second FRB-associated Magnetar X-ray Burst from SGR J1935+2154

Fast radio burst (FRB) is mysterious phenomenon with millisecond-duration radio pulses observed mostly from cosmological distance. The association between FRB 200428 and a magnetar X-ray burst (MXB) from SGR J1935+2154 has significantly advanced the understanding of FRB and magnetar bursts. However, it is uncertain whether this association between MXB and FRB (i.e. MXB/FRB 200428) is genuine or just coincidental only based on this single event. Here we report the discovery of a bright ($\rm\sim7.6\times10^{-7}\,erg \cdot cm^{-2}$ in 1-250 keV) magnetar X-ray burst detected by GECAM on October 14th, 2022 (dubbed as MXB 221014) from SGR J1935+2154, which is associated with a FRB detected by CHIME and GBT. We conducted a detailed temporal and spectral analysis of MXB 221014 with GECAM data and find that it is a bright and typical ($T_{90}\sim$250 ms) X-ray burst from this magnetar. Interestingly, we find two narrow X-ray pulses in the MXB, one of which temporally aligns with the main pulse of the FRB 221014 $\sim5.70$ ms latter than the peak time of FRB 221014), resembling the feature found in MXB/FRB 200428. Furthermore, we did comprehensive comparison between MXB/FRB 221014 and MXB/FRB 200428, and find that while the two events share several common features, they also exhibit distinct differences, highlighting the variety of the MXB-FRB association morphology. This finding not only confirms the association between MXB and FRB but also provides new insights into the mechanism of and the relationship between FRB and MXB.

astro-ph.HE↗
arXiv2026

DrugPlayGround: Benchmarking Large Language Models and Embeddings for Drug Discovery

Large language models (LLMs) are in the ascendancy for research in drug discovery, offering unprecedented opportunities to reshape drug research by accelerating hypothesis generation, optimizing candidate prioritization, and enabling more scalable and cost-effective drug discovery pipelines. However there is currently a lack of objective assessments of LLM performance to ascertain their advantages and limitations over traditional drug discovery platforms. To tackle this emergent problem, we have developed DrugPlayGround, a framework to evaluate and benchmark LLM performance for generating meaningful text-based descriptions of physiochemical drug characteristics, drug synergism, drug-protein interactions, and the physiological response to perturbations introduced by drug molecules. Moreover, DrugPlayGround is designed to work with domain experts to provide detailed explanations for justifying the predictions of LLMs, thereby testing LLMs for chemical and biological reasoning capabilities to push their greater use at the frontier of drug discovery at all of its stages.

cs.LG↗
arXiv2026

Systematic Study of the Simultaneous Events Detected by GECAM

GECAM is a constellation of all-sky monitors in hard X-ray and gamma-ray band primarily aimed at high energy transients such as gamma-ray bursts, soft gamma-ray repeaters, solar flares and terrestrial gamma-ray flashes. As GECAM has the highest temporal resolution (0.1~$μ$s) among instruments of its kind, it can identify the so-called simultaneous events (STE) that deposit signals in multiple detectors nearly at the same time (with a 0.3~$μ$s window). However, the properties and origin of STE have not yet been explored. In this work, we implemented, for the first time, a comprehensive analysis of the STE detected by GECAM, including their morphology, energy deposition, and the dependence on the geomagnetic coordinates. We find that these STE probably result from direct interactions between high-energy charged cosmic rays and satellite. These results demonstrate that GECAM can detect, identify, and characterize high-energy cosmic rays, making it a Micro Cosmic-Ray Observatory (MICRO) in low Earth orbit.

astro-ph.HE↗
arXiv2026

AffectGPT-R1: Leveraging Reinforcement Learning for Open-Vocabulary Multimodal Emotion Recognition

Open-Vocabulary Multimodal Emotion Recognition (OV-MER) aims to predict emotions without being constrained by label spaces, enabling fine-grained emotion understanding. Unlike traditional discriminative methods, OV-MER leverages generative models to capture the full spectrum of emotions and employs emotion wheels (EWs) for metric calculation. Previous approaches (e.g., AffectGPT) primarily rely on token-level loss during training. However, this objective is misaligned with the metrics used in OV-MER, while these metrics cannot be optimized via gradient backpropagation. To address this limitation, we propose AffectGPT-R1, a reinforcement learning framework that treats EW-based metrics as a reward function and applies policy optimization to maximize this reward. Additionally, we introduce an explicit reasoning process and examine its necessity in OV-MER. To further guide model behavior, we incorporate auxiliary rewards that regularize both emotion reasoning and emotion prediction. We also apply length penalties to mitigate reward hacking. Experimental results demonstrate that AffectGPT-R1 yields significant performance improvements on OV-MER. Moreover, our approach enhances generalized emotion understanding, achieving state-of-the-art results on MER-UniBench. Our code is provided in the supplementary material and will be released to facilitate future research.

cs.HC↗
arXiv2026

FreqLens: Interpretable Frequency Attribution for Time Series Forecasting

Time series forecasting models often lack interpretability, limiting their adoption in domains requiring explainable predictions. We propose \textsc{FreqLens}, an interpretable forecasting framework that discovers and attributes predictions to learnable frequency components. \textsc{FreqLens} introduces two key innovations: (1) \emph{learnable frequency discovery} -- frequency bases are parameterized via sigmoid mapping and learned from data with diversity regularization, enabling automatic discovery of dominant periodic patterns without domain knowledge; and (2) \emph{axiomatic frequency attribution} -- a theoretically grounded framework that provably satisfies Completeness, Faithfulness, Null-Frequency, and Symmetry axioms, with per-frequency attributions equivalent to Shapley values. On Traffic and Weather datasets, \textsc{FreqLens} achieves competitive or superior performance while discovering physically meaningful frequencies: all 5 independent runs discover the 24-hour daily cycle ($24.6 \pm 0.1$h, 2.5\% error) and 12-hour half-daily cycle ($11.8 \pm 0.1$h, 1.6\% error) on Traffic, and weekly cycles ($10\times$ longer than the input window) on Weather. These results demonstrate genuine frequency-level knowledge discovery with formal theoretical guarantees on attribution quality.

cs.LG↗
arXiv2026

Breaking Scale Anchoring: Frequency Representation Learning for Accurate High-Resolution Inference from Low-Resolution Training

Zero-Shot Super-Resolution Spatiotemporal Forecasting requires a deep learning model to be trained on low-resolution data and deployed for inference on high-resolution. Existing studies consider maintaining similar error across different resolutions as indicative of successful multi-resolution generalization. However, deep learning models serving as alternatives to numerical solvers should reduce error as resolution increases. The fundamental limitation is, the upper bound of physical law frequencies that low-resolution data can represent is constrained by its Nyquist frequency, making it difficult for models to process signals containing unseen frequency components during high-resolution inference. This results in errors being anchored at low resolution, incorrectly interpreted as successful generalization. We define this fundamental phenomenon as a new problem distinct from existing issues: Scale Anchoring. Therefore, we propose architecture-agnostic Frequency Representation Learning. It alleviates Scale Anchoring through resolution-aligned frequency representations and spectral consistency training: on grids with higher Nyquist frequencies, the frequency response in high-frequency bands of FRL-enhanced variants is more stable. This allows errors to decrease with resolution and significantly outperform baselines within our task and resolution range, while incurring only modest computational overhead.

cs.CV↗
arXiv2026

GeoLanG: Geometry-Aware Language-Guided Grasping with Unified RGB-D Multimodal Learning

Language-guided grasping has emerged as a promising paradigm for enabling robots to identify and manipulate target objects through natural language instructions, yet it remains highly challenging in cluttered or occluded scenes. Existing methods often rely on multi-stage pipelines that separate object perception and grasping, which leads to limited cross-modal fusion, redundant computation, and poor generalization in cluttered, occluded, or low-texture scenes. To address these limitations, we propose GeoLanG, an end-to-end multi-task framework built upon the CLIP architecture that unifies visual and linguistic inputs into a shared representation space for robust semantic alignment and improved generalization. To enhance target discrimination under occlusion and low-texture conditions, we explore a more effective use of depth information through the Depth-guided Geometric Module (DGGM), which converts depth into explicit geometric priors and injects them into the attention mechanism without additional computational overhead. In addition, we propose Adaptive Dense Channel Integration, which adaptively balances the contributions of multi-layer features to produce more discriminative and generalizable visual representations. Extensive experiments on the OCID-VLG dataset, as well as in both simulation and real-world hardware, demonstrate that GeoLanG enables precise and robust language-guided grasping in complex, cluttered environments, paving the way toward more reliable multimodal robotic manipulation in real-world human-centric settings.

cs.RO↗
arXiv2026

Accelerated Time-Domain Simulation of Complex Photonic Structures with a Data-Aware Fourier Neural Operator

Efficient and accurate time-domain simulation of electromagnetic fields in complex photonic devices is critical for designing broadband and ultrafast optical components, yet it is often limited by the high computational cost of conventional numerical methods like FDTD. While machine learning approaches show promise in accelerating these simulations, existing models still struggle to simultaneously capture the dynamic field evolution and generalize to complex geometries. In this paper, we introduce a Data-Aware Fourier Neural Operator (DA-FNO) as an innovative neural operator for solving electromagnetic simulations. Applied autoregressively, the model iteratively predicts the time-domain evolution of all field components and automatically terminates upon energy convergence. Our model not only generalizes to complex and randomized geometries but also shows good predictive consistency across the optical C-band (1530-1565nm) when evaluated on the test set. In a representative configuration, it achieves an 11* speedup over conventional methods while maintaining about 95% accuracy across the C-band. This approach provides a new pathway for C-band photonic simulations, potentially facilitating the research, development, and inverse design of novel photonic devices.

physics.optics↗