Yihao Hu

Yihao Hu, Zhihao Wen, Xiujin Liu et al.

Large Language Model (LLM) agents are increasingly improved through interaction, yet most self-evolution methods adapt either the policy or the learning environment in isolation. We identify this structural gap as \emph{Agent-Environment Misalignment}: the agent's capability frontier changes during training, while the environment that provides supervision remains static or only weakly coupled to the agent's revealed failures. We propose SEAL, a closed-loop co-evolution framework for interactive tool-use agents. SEAL collects on-policy trajectories under executable verification, diagnoses failed rollouts into turn-level failure labels, and uses these diagnoses as a shared signal for both environment-side adaptation and model-side policy optimization. The environment evolves its training-time learning interface by exposing clearer tool affordance cues, constraint information, and recovery-oriented feedback, while the policy is updated with diagnosis-guided advantage reweighting. Extensive experiments across in-distribution and out-of-distribution multi-turn tool-use evaluations show that SEAL improves low-resource agent learning: with only 400 training samples, it yields +8.25 to +26.25 average-point gains across three backbones and exhibits positive out-of-distribution transfer. These results demonstrate the value of jointly adapting the learner and its training-time learning substrate for robust self-improving LLM agents.

Pan Wang, Yihao Hu, Xiujin Liu et al.

Vision-language model (VLM) agents increasingly rely on memory-augmented reinforcement learning to reuse experience across long-horizon tasks, yet most existing frameworks store memory as text and depend on proprietary teacher models to summarize or refine it. This design is poorly matched to spatial decision making: geometric priors are compressed into lossy language, and sparse interaction is often supervised through delayed textual feedback rather than dense visually grounded signals. We argue that reusable experience for VLM agents should remain visually grounded. Based on this insight, we propose \textbf{AtlasVA}, a teacher-free visual skill memory framework that organizes memory into three complementary layers: spatial heatmaps, visual exemplars, and symbolic text skills. AtlasVA further evolves danger and affinity atlases directly from trajectory statistics and lightweight grid heuristics, and reuses these self-evolving atlases as potential-based shaping rewards for reinforcement learning. This unifies perception, memory, and optimization without external LLM supervision. Experiments on \textsc{Sokoban}, \textsc{FrozenLake}, 3D embodied navigation, and 3D robotic manipulation benchmarks show that AtlasVA consistently outperforms text-centric memory baselines and competitive VLM agents, with especially strong gains on spatially intensive tasks. Homepage: https://wangpan-ustc.github.io/AtlasvaWeb

Zhangquan Chen, Jiale Tao, Ruihuang Li et al.

While humans perceive the world through diverse modalities that operate synergistically to support a holistic understanding of their surroundings, existing omnivideo models still face substantial challenges on audio-visual understanding tasks. In this paper, we propose OmniVideo-R1, a novel reinforced framework that improves mixed-modality reasoning. OmniVideo-R1 empowers models to "think with omnimodal cues" by two key strategies: (1) query-intensive grounding based on self-supervised learning paradigms; and (2) modality-attentive fusion built upon contrastive learning paradigms. Extensive experiments on multiple benchmarks demonstrate that OmniVideo-R1 consistently outperforms strong baselines, highlighting its effectiveness and robust generalization capabilities.

Pan Wang, Yihao Hu, Xiujin Liu et al.

Traditional shadow removal networks often treat image restoration as an unconstrained mapping, lacking the physical interpretability required to balance localized texture recovery with global illumination consistency. To address this, we propose CFSR, a multi-modal prior-driven framework that reframes shadow removal as a physics-constrained restoration process. By seamlessly integrating 3D geometric cues with large-scale foundation model semantics, CFSR effectively bridges the 2D-3D domain gap. Specifically, we first map observations into a custom HVI color space to suppress shadow-induced noise and robustly fuse RGB data with estimated depth priors. At its core, our Geometric & Semantic Dual Explicit Guided Attention mechanism utilizes DINO features and 3D surface normals to directly modulate the attention affinity matrix, structurally enforcing physical lighting constraints. To recover severely degraded regions, we inject holistic priors via a frozen CLIP encoder. Finally, our Frequency Collaborative Reconstruction Module (FCRM) achieves an optimal synthesis by decoupling the decoding process. Conditioned on geometric priors, FCRM seamlessly harmonizes the reconstruction of sharp high-frequency occlusion boundaries with the restoration of low-frequency global illumination. Extensive experiments demonstrate that CFSR achieves state-of-the-art performance across multiple challenging benchmarks.

Pan Wang, Yihao Hu, Xiaodong Bai et al.

As a specialty agricultural product with a large market scale, Shatian pomelo necessitates the adoption of automated detection to ensure accurate quantity and meet commercial demands for lean production. Existing research often involves specialized networks tailored for specific theoretical or dataset scenarios, but these methods tend to degrade performance in real-world. Through analysis of factors in this issue, this study identifies four key challenges that affect the accuracy of Shatian pomelo detection: imaging devices, lighting conditions, object scale variation, and occlusion. To mitigate these challenges, a multi-strategy framework is proposed in this paper. Firstly, to effectively solve tone variation introduced by diverse imaging devices and complex orchard environments, we utilize a multi-scenario dataset, STP-AgriData, which is constructed by integrating real orchard images with internet-sourced data. Secondly, to simulate the inconsistent illumination conditions, specific data augmentations such as adjusting contrast and changing brightness, are applied to the above dataset. Thirdly, to address the issues of object scale variation and occlusion in fruit detection, an REAS-Det network is designed in this paper. For scale variation, RFAConv and C3RFEM modules are designed to expand and enhance the receptive fields. For occlusion variation, a multi-scale, multi-head feature selection structure (MultiSEAM) and soft-NMS are introduced to enhance the handling of occlusion issues to improve detection accuracy. The results of these experiments achieved a precision(P) of 87.6%, a recall (R) of 74.9%, a mAP@.50 of 82.8%, and a mAP@.50:.95 of 53.3%. Our proposed network demonstrates superior performance compared to other state-of-the-art detection methods.

Yihao Hu, Pan Wang, Xiaodong Bai et al.

Pomelo detection is an essential process for their localization, automated robotic harvesting, and maturity analysis. However, detecting Shatian pomelo in complex orchard environments poses significant challenges, including multi-scale issues, obstructions from trunks and leaves, small object detection, etc. To address these issues, this study constructs a custom dataset STP-AgriData and proposes the SDE-DET model for Shatian pomelo detection. SDE-DET first utilizes the Star Block to effectively acquire high-dimensional information without increasing the computational overhead. Furthermore, the presented model adopts Deformable Attention in its backbone, to enhance its ability to detect pomelos under occluded conditions. Finally, multiple Efficient Multi-Scale Attention mechanisms are integrated into our model to reduce the computational overhead and extract deep visual representations, thereby improving the capacity for small object detection. In the experiment, we compared SDE-DET with the Yolo series and other mainstream detection models in Shatian pomelo detection. The presented SDE-DET model achieved scores of 0.883, 0.771, 0.838, 0.497, and 0.823 in Precision, Recall, mAP@0.5, mAP@0.5:0.95 and F1-score, respectively. SDE-DET has achieved state-of-the-art performance on the STP-AgriData dataset. Experiments indicate that the SDE-DET provides a reliable method for Shatian pomelo detection, laying the foundation for the further development of automatic harvest robots.

Congyu Qiao, Ning Xu, Yihao Hu et al.

Instance-dependent Partial Label Learning (ID-PLL) aims to learn a multi-class predictive model given training instances annotated with candidate labels related to features, among which correct labels are hidden fixed but unknown. The previous works involve leveraging the identification capability of the training model itself to iteratively refine supervision information. However, these methods overlook a critical aspect of ID-PLL: the training model is prone to overfitting on incorrect candidate labels, thereby providing poor supervision information and creating a bottleneck in training. In this paper, we propose to leverage reduction-based pseudo-labels to alleviate the influence of incorrect candidate labels and train our predictive model to overcome this bottleneck. Specifically, reduction-based pseudo-labels are generated by performing weighted aggregation on the outputs of a multi-branch auxiliary model, with each branch trained in a label subspace that excludes certain labels. This approach ensures that each branch explicitly avoids the disturbance of the excluded labels, allowing the pseudo-labels provided for instances troubled by these excluded labels to benefit from the unaffected branches. Theoretically, we demonstrate that reduction-based pseudo-labels exhibit greater consistency with the Bayes optimal classifier compared to pseudo-labels directly generated from the predictive model.

Yihao Hu, Fearghal O'Donncha, Paulito Palmes et al.

This paper presents a novel spatio-temporal LSTM (SPATIAL) architecture for time series forecasting applied to environmental datasets. The framework was evaluated across multiple sensors and for three different oceanic variables: current speed, temperature, and dissolved oxygen. Network implementation proceeded in two directions that are nominally separated but connected as part of a natural environmental system -- across the spatial (between individual sensors) and temporal components of the sensor data. Data from four sensors sampling current speed, and eight measuring both temperature and dissolved oxygen evaluated the framework. Results were compared against RF and XGB baseline models that learned on the temporal signal of each sensor independently by extracting the date-time features together with the past history of data using sliding window matrix. Results demonstrated ability to accurately replicate complex signals and provide comparable performance to state-of-the-art benchmarks. Notably, the novel framework provided a simpler pre-processing and training pipeline that handles missing values via a simple masking layer. Enabling learning across the spatial and temporal directions, this paper addresses two fundamental challenges of ML applications to environmental science: 1) data sparsity and the challenges and costs of collecting measurements of environmental conditions such as ocean dynamics, and 2) environmental datasets are inherently connected in the spatial and temporal directions while classical ML approaches only consider one of these directions. Furthermore, sharing of parameters across all input steps makes SPATIAL a fast, scalable, and easily-parameterized forecasting framework.

Yihao Hu, Tong Zhao, Shixin Xu et al.

Partial differential equations (PDEs) play a crucial role in studying a vast number of problems in science and engineering. Numerically solving nonlinear and/or high-dimensional PDEs is often a challenging task. Inspired by the traditional finite difference and finite elements methods and emerging advancements in machine learning, we propose a sequence deep learning framework called Neural-PDE, which allows to automatically learn governing rules of any time-dependent PDE system from existing data by using a bidirectional LSTM encoder, and predict the next n time steps data. One critical feature of our proposed framework is that the Neural-PDE is able to simultaneously learn and simulate the multiscale variables.We test the Neural-PDE by a range of examples from one-dimensional PDEs to a high-dimensional and nonlinear complex fluids model. The results show that the Neural-PDE is capable of learning the initial conditions, boundary conditions and differential operators without the knowledge of the specific form of a PDE system.In our experiments the Neural-PDE can efficiently extract the dynamics within 20 epochs training, and produces accurate predictions. Furthermore, unlike the traditional machine learning approaches in learning PDE such as CNN and MLP which require vast parameters for model precision, Neural-PDE shares parameters across all time steps, thus considerably reduces the computational complexity and leads to a fast learning algorithm.

Yihao Hu, Ari Trachtenberg, Prakash Ishwar

Real-time, online-editing web apps provide free and convenient services for collaboratively editing, sharing and storing files. The benefits of these web applications do not come for free: not only do service providers have full access to the users' files, but they also control access, transmission, and storage mechanisms for them. As a result, user data may be at risk of data mining, third-party interception, or even manipulation. To combat this, we propose a new system for helping to preserve the privacy of user data within collaborative environments. There are several distinct challenges in producing such a system, including developing an encryption mechanism that does not interfere with the back-end (and often proprietary) control mechanisms utilized by the service, and identifying transparent code hooks through which to obfuscate user data. Toward the first challenge, we develop a character-level encryption scheme that is more resilient to the types of attacks that plague classical substitution ciphers. For the second challenge, we design a browser extension that robustly demonstrates the feasibility of our approach, and show a concrete implementation for Google Chrome and the widely-used Google Docs platform. Our example tangibly demonstrates how several users with a shared key can collaboratively and transparently edit a Google Docs document without revealing the plaintext directly to Google.