Hongxiang Lin

Hongxiang Lin, Matteo Figini, Felice D'Arco et al.

Low-field (<1T) magnetic resonance imaging (MRI) scanners remain in widespread use in low- and middle-income countries (LMICs) and are commonly used for some applications in higher income countries e.g. for small child patients with obesity, claustrophobia, implants, or tattoos. However, low-field MR images commonly have lower resolution and poorer contrast than images from high field (1.5T, 3T, and above). Here, we present Image Quality Transfer (IQT) to enhance low-field structural MRI by estimating from a low-field image the image we would have obtained from the same subject at high field. Our approach uses (i) a stochastic low-field image simulator as the forward model to capture uncertainty and variation in the contrast of low-field images corresponding to a particular high-field image, and (ii) an anisotropic U-Net variant specifically designed for the IQT inverse problem. We evaluate the proposed algorithm both in simulation and using multi-contrast (T1-weighted, T2-weighted, and fluid attenuated inversion recovery (FLAIR)) clinical low-field MRI data from an LMIC hospital. We show the efficacy of IQT in improving contrast and resolution of low-field MR images. We demonstrate that IQT-enhanced images have potential for enhancing visualisation of anatomical structures and pathological lesions of clinical relevance from the perspective of radiologists. IQT is proved to have capability of boosting the diagnostic value of low-field MRI, especially in low-resource settings.

Stefano B. Blumberg, Hongxiang Lin, Francesco Grussu et al.

We present PROSUB: PROgressive SUBsampling, a deep learning based, automated methodology that subsamples an oversampled data set (e.g. multi-channeled 3D images) with minimal loss of information. We build upon a recent dual-network approach that won the MICCAI MUlti-DIffusion (MUDI) quantitative MRI measurement sampling-reconstruction challenge, but suffers from deep learning training instability, by subsampling with a hard decision boundary. PROSUB uses the paradigm of recursive feature elimination (RFE) and progressively subsamples measurements during deep learning training, improving optimization stability. PROSUB also integrates a neural architecture search (NAS) paradigm, allowing the network architecture hyperparameters to respond to the subsampling process. We show PROSUB outperforms the winner of the MUDI MICCAI challenge, producing large improvements >18% MSE on the MUDI challenge sub-tasks and qualitative improvements on downstream processes useful for clinical applications. We also show the benefits of incorporating NAS and analyze the effect of PROSUB's components. As our method generalizes to other problems beyond MRI measurement selection-reconstruction, our code is https://github.com/sbb-gh/PROSUB

Hongxiang Lin, Zhirui Kuai, Erpeng Xue et al.

Structured skill prompts improve exploration in long-horizon agentic reinforcement learning (RL). Skill-augmented RL methods retain external skills at inference, while skill-internalization RL methods withdraw them during training to enable autonomous performance. However, existing internalization approaches only use skill-helpfulness contrast for curriculum control, leaving the policy update unchanged and unable to distinguish skill-dependent from autonomous success. We propose SkillC, a framework based on Contrastive Skill Credit Assignment (CSCA) that converts this contrast into a direct learning signal for internalization. \textsc{SkillC} samples paired skill-injected and skill-free rollouts for tasks from active skill types within the same policy update, and injects their task-level contrast into optimization via a dual-stream advantage estimator that preserves global ranking while applying a one-sided correction toward skill-free success. A smoothed validation-level signal further drives an adaptive curriculum over attribution strength, rollout allocation, and monotonic active-set pruning. Experiments on ALFWorld and WebShop show that, without runtime skill access, SkillC surpasses the strongest prior skill-internalization RL baseline by 5.5\% and 4.4\%, respectively, while remaining competitive with skill-augmented RL methods.

Minghui Wu, Yangdi Xu, Yingying Xu et al.

Large Vision Model (LVM) has recently demonstrated great potential for medical imaging tasks, potentially enabling image enhancement for sparse-view Cone-Beam Computed Tomography (CBCT), despite requiring a substantial amount of data for training. Meanwhile, Deep Image Prior (DIP) effectively guides an untrained neural network to generate high-quality CBCT images without any training data. However, the original DIP method relies on a well-defined forward model and a large-capacity backbone network, which is notoriously difficult to converge. In this paper, we propose a stable optimization method for the forward-model-free, LVM-based DIP model for sparse-view CBCT. Our approach consists of two main characteristics: (1) multi-scale perceptual loss (MSPL) which measures the similarity of perceptual features between the reference and output images at multiple resolutions without the need for any forward model, and (2) a reweighting mechanism that stabilizes the iteration trajectory of MSPL. One shot optimization is used to simultaneously and stably reweight MSPL and optimize LVM. We evaluate our approach on two publicly available datasets: SPARE and Walnut. The results show significant improvements in both image quality metrics and visualization that demonstrates reduced streak artifacts. The source code is available upon request.

Hongxiang Lin, Zhirui Kuai, Erpeng Xue et al.

Test-time reinforcement learning (TTRL) reports substantial accuracy gains on mathematical reasoning benchmarks using majority vote as a pseudo-label signal. We argue these gains are systematically misinterpreted: most reflect sharpening of already-solvable problems rather than genuine learning, while problems corrupted from correct to incorrect outnumber truly learned ones, and this damage is irreversible once majority vote locks onto a wrong answer. Per-problem tracking reveals that correct-answer signals in low-ability problems are briefly active before being permanently suppressed, a phenomenon we term the \textit{Correct-Answer Extinction Window}, with Flip Rate (FR) as its leading indicator. We thus propose \textbf{TTRL-Guard}, a lightweight framework with three mechanisms targeting the extinction window: Flip-Rate-Aware Reward Scaling (FRS) down-weights at-risk updates as FR declines, Minority-Preserving Sampling (MPS) retains gradient signal from minority correct answers, and Risk-Conditioned Sparse Updatings (RCSU) suspends updates on polarized problems. Experiments across three models and four benchmarks show that TTRL-Guard achieves the best average pass@1 on Qwen2.5-7B-Instruct and Qwen3-4B, improves relatively over TTRL by +54\% on AIME 2025. \footnote{Our code and implementation details are available at https://github.com/linhxkkkk/TTRL-Guard.

Hongxiang Lin, Takashi Azuma, Mehmet Burcin Unlu et al.

Photo-Acoustic Tomography (PAT) can reconstruct a distribution of optical absorbers acting as instantaneous sound sources in subcutaneous microvasculature of a human breast. Adjoint methods for PAT, typically Time-Reversal (TR) and Back-Projection (BP), are ways to refocus time-reversed acoustic signals on sources by wave propagation from the position of sensors. TR and BP have different treatments for received signals, but they are equivalent under continuously sampling on a closed circular sensor array in two dimensions. Here, we analyze image quality with discrete under-sampled sensors in the sense of the Shannon sampling theorem. We investigate resolution and contrast of TR and BP, respectively in one source-sensor pair configuration and the frequency domain. With Hankel's asymptotic expansion to the integrands of imaging functions, our main contribution is to demonstrate that TR and BP have better performance on contrast and resolution, respectively. We also show that the integrand of TR includes additional side lobes which degrade axial resolution whereas that of BP conversely has relatively small amplitudes. Moreover, omnidirectional resolution is improved if more sensors are employed to collect the received signals. Nevertheless, for the under-sampled sensors, we propose the Truncated Back-Projection (TBP) method to enhance the contrast of BP using removing higher frequency components in the received signals. We conduct numerical experiments on the two-dimensional projected phantom model extracted from OA-Breast Database. The experiments verify our theories and show that the proposed TBP possesses better omnidirectional resolution as well as contrast compared with TR and BP with under-sampled sensors.

Zhiwei Lin, Yongtao Wang, Hongxiang Lin

Recently, self-supervised representation learning gives further development in multimedia technology. Most existing self-supervised learning methods are applicable to packaged data. However, when it comes to streamed data, they are suffering from a catastrophic forgetting problem, which is not studied extensively. In this paper, we make the first attempt to tackle the catastrophic forgetting problem in the mainstream self-supervised methods, i.e., contrastive learning methods. Specifically, we first develop a rehearsal-based framework combined with a novel sampling strategy and a self-supervised knowledge distillation to transfer information over time efficiently. Then, we propose an extra sample queue to help the network separate the feature representations of old and new data in the embedding space. Experimental results show that compared with the naive self-supervised baseline, which learns tasks one by one without taking any technique, we improve the image classification accuracy by 1.60% on CIFAR-100, 2.86% on ImageNet-Sub, and 1.29% on ImageNet-Full under 10 incremental steps setting. Our code will be available at https://github.com/VDIGPKU/ContinualContrastiveLearning.

Yukun Zhou, Moucheng Xu, Yipeng Hu et al.

Accurate multi-class segmentation is a long-standing challenge in medical imaging, especially in scenarios where classes share strong similarity. Segmenting retinal blood vessels in retinal photographs is one such scenario, in which arteries and veins need to be identified and differentiated from each other and from the background. Intra-segment misclassification, i.e. veins classified as arteries or vice versa, frequently occurs when arteries and veins intersect, whereas in binary retinal vessel segmentation, error rates are much lower. We thus propose a new approach that decomposes multi-class segmentation into multiple binary, followed by a binary-to-multi-class fusion network. The network merges representations of artery, vein, and multi-class feature maps, each of which are supervised by expert vessel annotation in adversarial training. A skip-connection based merging process explicitly maintains class-specific gradients to avoid gradient vanishing in deep layers, to favor the discriminative features. The results show that, our model respectively improves F1-score by 4.4\%, 5.1\%, and 4.2\% compared with three state-of-the-art deep learning based methods on DRIVE-AV, LES-AV, and HRF-AV data sets. Code: https://github.com/rmaphoh/Learning-AVSegmentation

Zhaoyu Hu, Jianyang Wang, Hao Guo et al.

In the context of the booming digital economy, recommendation systems, as a key link connecting users and numerous services, face challenges in modeling user behavior sequences on local-life service platforms, including the sparsity of long sequences and strong spatio-temporal dependence. Such challenges can be addressed by drawing an analogy to the forgetting process in human memory. This is because users' responses to recommended content follow the recency effect and the cyclicality of memory. By exploring this, this paper introduces the forgetting curve and proposes Spatio-Temporal periodic Interest Modeling (STIM) with long sequences for local-life service recommendation. STIM integrates three key components: a dynamic masking module based on the forgetting curve, which is used to extract both recent spatiotemporal features and periodic spatiotemporal features; a query-based mixture of experts (MoE) approach that can adaptively activate expert networks under different dynamic masks, enabling the collaborative modeling of time, location, and items; and a hierarchical multi-interest network unit, which captures multi-interest representations by modeling the hierarchical interactions between the shallow and deep semantics of users' recent behaviors. By introducing the STIM method, we conducted online A/B tests and achieved a 1.54\% improvement in gross transaction volume (GTV). In addition, extended offline experiments also showed improvements. STIM has been deployed in a large-scale local-life service recommendation system, serving hundreds of millions of daily active users in core application scenarios.

Xianyang Qi, Yuan Tian, Zhaoyu Hu et al.

Industrial recommender systems critically depend on high-quality ranking models. However, traditional pipelines still rely on manual feature engineering and scenario-specific architectures, which hinder cross-scenario transfer and large-scale deployment. To address these challenges, we propose \textbf{MTmixAtt}, a unified Mixture-of-Experts (MoE) architecture with Multi-Mix Attention, designed for large-scale recommendation tasks. MTmixAtt integrates two key components. The \textbf{AutoToken} module automatically clusters heterogeneous features into semantically coherent tokens, removing the need for human-defined feature groups. The \textbf{MTmixAttBlock} module enables efficient token interaction via a learnable mixing matrix, shared dense experts, and scenario-aware sparse experts, capturing both global patterns and scenario-specific behaviors within a single framework. Extensive experiments on the industrial TRec dataset from Meituan demonstrate that MTmixAtt consistently outperforms state-of-the-art baselines including Transformer-based models, WuKong, HiFormer, MLP-Mixer, and RankMixer. At comparable parameter scales, MTmixAtt achieves superior CTR and CTCVR metrics; scaling to MTmixAtt-1B yields further monotonic gains. Large-scale online A/B tests validate the real-world impact: in the \textit{Homepage} scenario, MTmixAtt increases Payment PV by \textbf{+3.62\%} and Actual Payment GTV by \textbf{+2.54\%}. Overall, MTmixAtt provides a unified and scalable solution for modeling arbitrary heterogeneous features across scenarios, significantly improving both user experience and commercial outcomes.

Hongxiang Lin, Hao Guo, Zeshun Li et al.

Traditional recommendation systems tend to trap users in strong feedback loops by excessively pushing content aligned with their historical preferences, thereby limiting exploration opportunities and causing content fatigue. Although large language models (LLMs) demonstrate potential with their diverse content generation capabilities, existing LLM-enhanced dual-model frameworks face two major limitations: first, they overlook long-term preferences driven by group identity, leading to biased interest modeling; second, they suffer from static optimization flaws, as a one-time alignment process fails to leverage incremental user data for closed-loop optimization. To address these challenges, we propose the Co-Evolutionary Alignment (CoEA) method. For interest modeling bias, we introduce Dual-Stable Interest Exploration (DSIE) module, jointly modeling long-term group identity and short-term individual interests through parallel processing of behavioral sequences. For static optimization limitations, we design a Periodic Collaborative Optimization (PCO) mechanism. This mechanism regularly conducts preference verification on incremental data using the Relevance LLM, then guides the Novelty LLM to perform fine-tuning based on the verification results, and subsequently feeds back the output of the incrementally fine-tuned Novelty LLM to the Relevance LLM for re-evaluation, thereby achieving a dynamic closed-loop optimization. Extensive online and offline experiments verify the effectiveness of the CoEA model in exploratory recommendation.

Matteo Figini, Hongxiang Lin, Godwin Ogbole et al.

1.5T or 3T scanners are the current standard for clinical MRI, but low-field (<1T) scanners are still common in many lower- and middle-income countries for reasons of cost and robustness to power failures. Compared to modern high-field scanners, low-field scanners provide images with lower signal-to-noise ratio at equivalent resolution, leaving practitioners to compensate by using large slice thickness and incomplete spatial coverage. Furthermore, the contrast between different types of brain tissue may be substantially reduced even at equal signal-to-noise ratio, which limits diagnostic value. Recently the paradigm of Image Quality Transfer has been applied to enhance 0.36T structural images aiming to approximate the resolution, spatial coverage, and contrast of typical 1.5T or 3T images. A variant of the neural network U-Net was trained using low-field images simulated from the publicly available 3T Human Connectome Project dataset. Here we present qualitative results from real and simulated clinical low-field brain images showing the potential value of IQT to enhance the clinical utility of readily accessible low-field MRIs in the management of epilepsy.

Hongxiang Lin, Matteo Figini, Ryutaro Tanno et al.

MR images scanned at low magnetic field ($<1$T) have lower resolution in the slice direction and lower contrast, due to a relatively small signal-to-noise ratio (SNR) than those from high field (typically 1.5T and 3T). We adapt the recent idea of Image Quality Transfer (IQT) to enhance very low-field structural images aiming to estimate the resolution, spatial coverage, and contrast of high-field images. Analogous to many learning-based image enhancement techniques, IQT generates training data from high-field scans alone by simulating low-field images through a pre-defined decimation model. However, the ground truth decimation model is not well-known in practice, and lack of its specification can bias the trained model, aggravating performance on the real low-field scans. In this paper we propose a probabilistic decimation simulator to improve robustness of model training. It is used to generate and augment various low-field images whose parameters are random variables and sampled from an empirical distribution related to tissue-specific SNR on a 0.36T scanner. The probabilistic decimation simulator is model-agnostic, that is, it can be used with any super-resolution networks. Furthermore we propose a variant of U-Net architecture to improve its learning performance. We show promising qualitative results from clinical low-field images confirming the strong efficacy of IQT in an important new application area: epilepsy diagnosis in sub-Saharan Africa where only low-field scanners are normally available.