Reasoning in the Dark: Interleaved Vision-Text Reasoning in Latent Space

Multimodal reasoning aims to enhance the capabilities of MLLMs by incorporating intermediate reasoning steps before reaching the final answer. It has evolved from text-only reasoning to the integration of visual information, enabling the thought process to be conveyed through both images and text. Despite its effectiveness, current multimodal reasoning methods depend on explicit reasoning steps that require labor-intensive vision-text annotations and inherently introduce significant inference latency. To address these issues, we introduce multimodal latent reasoning with the advantages of multimodal representation, reduced annotation, and inference efficiency. To facilicate it, we propose Interleaved Vision-Text Latent Reasoning (IVT-LR), which injects both visual and textual information in the reasoning process within the latent space. Specifically, IVT-LR represents each reasoning step by combining two implicit parts: latent text (the hidden states from the previous step) and latent vision (a set of selected image embeddings). We further introduce a progressive multi-stage training strategy to enable MLLMs to perform the above multimodal latent reasoning steps. Experiments on M3CoT and ScienceQA demonstrate that our IVT-LR method achieves an average performance increase of 5.45% in accuracy, while simultaneously achieving a speed increase of over 5 times compared to existing approaches. Code available at https://github.com/FYYDCC/IVT-LR.

Flexible and Efficient Spatio-Temporal Transformer for Sequential Visual Place Recognition

Sequential Visual Place Recognition (Seq-VPR) leverages transformers to capture spatio-temporal features effectively; however, existing approaches prioritize performance at the expense of flexibility and efficiency. In practice, a transformer-based Seq-VPR model should be flexible to the number of frames per sequence (seq-length), deliver fast inference, and have low memory usage to meet real-time constraints. To our knowledge, no existing transformer-based Seq-VPR method achieves both flexibility and efficiency. To address this gap, we propose Adapt-STformer, a Seq-VPR method built around our novel Recurrent Deformable Transformer Encoder (Recurrent-DTE), which uses an iterative recurrent mechanism to fuse information from multiple sequential frames. This design naturally supports variable seq-lengths, fast inference, and low memory usage. Experiments on the Nordland, Oxford, and NuScenes datasets show that Adapt-STformer boosts recall by up to 17% while reducing sequence extraction time by 36% and lowering memory usage by 35% compared to the second-best baseline.

MATCH: Multi-faceted Adaptive Topo-Consistency for Semi-Supervised Histopathology Segmentation

In semi-supervised segmentation, capturing meaningful semantic structures from unlabeled data is essential. This is particularly challenging in histopathology image analysis, where objects are densely distributed. To address this issue, we propose a semi-supervised segmentation framework designed to robustly identify and preserve relevant topological features. Our method leverages multiple perturbed predictions obtained through stochastic dropouts and temporal training snapshots, enforcing topological consistency across these varied outputs. This consistency mechanism helps distinguish biologically meaningful structures from transient and noisy artifacts. A key challenge in this process is to accurately match the corresponding topological features across the predictions in the absence of ground truth. To overcome this, we introduce a novel matching strategy that integrates spatial overlap with global structural alignment, minimizing discrepancies among predictions. Extensive experiments demonstrate that our approach effectively reduces topological errors, resulting in more robust and accurate segmentations essential for reliable downstream analysis. Code is available at \href{https://github.com/Melon-Xu/MATCH}{https://github.com/Melon-Xu/MATCH}.

Bézier Meets Diffusion: Robust Generation Across Domains for Medical Image Segmentation

Training robust learning algorithms across different medical imaging modalities is challenging due to the large domain gap. Unsupervised domain adaptation (UDA) mitigates this problem by using annotated images from the source domain and unlabeled images from the target domain to train the deep models. Existing approaches often rely on GAN-based style transfer, but these methods struggle to capture cross-domain mappings in regions with high variability. In this paper, we propose a unified framework, B\'ezier Meets Diffusion, for cross-domain image generation. First, we introduce a B\'ezier-curve-based style transfer strategy that effectively reduces the domain gap between source and target domains. The transferred source images enable the training of a more robust segmentation model across domains. Thereafter, using pseudo-labels generated by this segmentation model on the target domain, we train a conditional diffusion model (CDM) to synthesize high-quality, labeled target-domain images. To mitigate the impact of noisy pseudo-labels, we further develop an uncertainty-guided score matching method that improves the robustness of CDM training. Extensive experiments on public datasets demonstrate that our approach generates realistic labeled images, significantly augmenting the target domain and improving segmentation performance.

Track Any Motions under Any Disturbances

A foundational humanoid motion tracker is expected to be able to track diverse, highly dynamic, and contact-rich motions. More importantly, it needs to operate stably in real-world scenarios against various dynamics disturbances, including terrains, external forces, and physical property changes for general practical use. To achieve this goal, we propose Any2Track (Track Any motions under Any disturbances), a two-stage RL framework to track various motions under multiple disturbances in the real world. Any2Track reformulates dynamics adaptability as an additional capability on top of basic action execution and consists of two key components: AnyTracker and AnyAdapter. AnyTracker is a general motion tracker with a series of careful designs to track various motions within a single policy. AnyAdapter is a history-informed adaptation module that endows the tracker with online dynamics adaptability to overcome the sim2real gap and multiple real-world disturbances. We deploy Any2Track on Unitree G1 hardware and achieve a successful sim2real transfer in a zero-shot manner. Any2Track performs exceptionally well in tracking various motions under multiple real-world disturbances.

Comprehensive Comparison Network: a framework for locality-aware, routes-comparable and interpretable route recommendation

Route recommendation (RR) is a core task of route planning in the Amap app, with the goal of recommending the optimal route among candidate routes to users. Unlike traditional recommendation methods, insights into the local quality of routes and comparisons between candidate routes are crucial for enhancing recommendation performance but often overlooked in previous studies. To achieve these, we propose a novel model called Comprehensive Comparison Network (CCN). CCN not only uses query-level features (e.g. user features) and item-level features (e.g. route features, item embedding) that are common in traditional recommendations, but also introduces comparison-level features which describe the non-overlapping segments between different routes to capture the local quality of routes. The key component Comprehensive Comparison Block (CCB) in CCN is designed to enable comparisons between routes. CCB includes a Comprehensive Comparison Operator (CCO) and a multi-scenario MLP, which can update the representations of candidate routes based on a comprehensive comparison. By stacking multiple CCBs, CCN can determine the final scores of candidate routes and recommend the optimal one to the user. Additionally, since routes directly affect the costs and risks experienced by users, the RR model must be interpretable for online deployment. Therefore, we designed an interpretable pair scoring network to achieve interpretability. Both offline and online experiments demonstrate that CCN significantly improves RR performance and exhibits strong interpretability. CCN has been fully deployed in the Amap app for over a year, providing stable and optimal benefits for route recommendations.

ImmunoDiff: A Diffusion Model for Immunotherapy Response Prediction in Lung Cancer

Accurately predicting immunotherapy response in Non-Small Cell Lung Cancer (NSCLC) remains a critical unmet need. Existing radiomics and deep learning-based predictive models rely primarily on pre-treatment imaging to predict categorical response outcomes, limiting their ability to capture the complex morphological and textural transformations induced by immunotherapy. This study introduces ImmunoDiff, an anatomy-aware diffusion model designed to synthesize post-treatment CT scans from baseline imaging while incorporating clinically relevant constraints. The proposed framework integrates anatomical priors, specifically lobar and vascular structures, to enhance fidelity in CT synthesis. Additionally, we introduce a novel cbi-Adapter, a conditioning module that ensures pairwise-consistent multimodal integration of imaging and clinical data embeddings, to refine the generative process. Additionally, a clinical variable conditioning mechanism is introduced, leveraging demographic data, blood-based biomarkers, and PD-L1 expression to refine the generative process. Evaluations on an in-house NSCLC cohort treated with immune checkpoint inhibitors demonstrate a 21.24% improvement in balanced accuracy for response prediction and a 0.03 increase in c-index for survival prediction. Code will be released soon.

Unleashing Humanoid Reaching Potential via Real-world-Ready Skill Space

Humans possess a large reachable space in the 3D world, enabling interaction with objects at varying heights and distances. However, realizing such large-space reaching on humanoids is a complex whole-body control problem and requires the robot to master diverse skills simultaneously-including base positioning and reorientation, height and body posture adjustments, and end-effector pose control. Learning from scratch often leads to optimization difficulty and poor sim2real transferability. To address this challenge, we propose Real-world-Ready Skill Space (R2S2). Our approach begins with a carefully designed skill library consisting of real-world-ready primitive skills. We ensure optimal performance and robust sim2real transfer through individual skill tuning and sim2real evaluation. These skills are then ensembled into a unified latent space, serving as a structured prior that helps task execution in an efficient and sim2real transferable manner. A high-level planner, trained to sample skills from this space, enables the robot to accomplish real-world goal-reaching tasks. We demonstrate zero-shot sim2real transfer and validate R2S2 in multiple challenging goal-reaching scenarios.

NTIRE 2025 Challenge on Cross-Domain Few-Shot Object Detection: Methods and Results

Cross-Domain Few-Shot Object Detection (CD-FSOD) poses significant challenges to existing object detection and few-shot detection models when applied across domains. In conjunction with NTIRE 2025, we organized the 1st CD-FSOD Challenge, aiming to advance the performance of current object detectors on entirely novel target domains with only limited labeled data. The challenge attracted 152 registered participants, received submissions from 42 teams, and concluded with 13 teams making valid final submissions. Participants approached the task from diverse perspectives, proposing novel models that achieved new state-of-the-art (SOTA) results under both open-source and closed-source settings. In this report, we present an overview of the 1st NTIRE 2025 CD-FSOD Challenge, highlighting the proposed solutions and summarizing the results submitted by the participants.

Constants of motion network revisited

Discovering constants of motion is meaningful in helping understand the dynamical systems, but inevitably needs proficient mathematical skills and keen analytical capabilities. With the prevalence of deep learning, methods employing neural networks, such as Constant Of Motion nETwork (COMET), are promising in handling this scientific problem. Although the COMET method can produce better predictions on dynamics by exploiting the discovered constants of motion, there is still plenty of room to sharpen it. In this paper, we propose a novel neural network architecture, built using the singular-value-decomposition (SVD) technique, and a two-phase training algorithm to improve the performance of COMET. Extensive experiments show that our approach not only retains the advantages of COMET, such as applying to non-Hamiltonian systems and indicating the number of constants of motion, but also can be more lightweight and noise-robust than COMET.