Chenyang Wu

Dehong Kong, Lina Lei, Lingtao Zheng et al.

Prior work on aesthetic composition typically produces a single aesthetically pleasing crop, overlooking the narrative value of composing multiple shots from one scene. In practice, multi-shot composition is critical for downstream creative workflows: commercial posters often require multiple crops with different emphases (e.g., context, subject, and emotion/product details) to present key story beats. Therefore, we propose \textbf{Triple-Shot Compositions (TSC)}, a composition task that generates a three-shot set -- establishing, medium, and close-up -- from a single human-centric image, each paired with a brief shot description to support visual narration. To learn TSC with limited expert annotations, we introduce \textbf{ShotCrop} which undergoes a three-stage training process: it first applies Chain-of-Thought supervised fine-tuning to establish basic reasoning and aesthetic shot-cropping skills, then performs semi-supervised fine-tuning with high-confidence pseudo labels to further enhance aesthetic capability, and is finally optimized with Group Relative Policy Optimization for \textbf{ShotCrop} (GRPO-S) using a composite reward tailored for it. Specifically, our pseudo-labeling strategy combines MLLM-based scoring, aesthetic assessment, and CLIP similarity to retain high-confidence training signals. In addition, we present TSC-Bench, a benchmark of 1.2k expert-annotated test cases. Notably, ShotCrop achieves an average improvement of \textbf{2.82} times over GPT-5 in shot localization accuracy.

Chen-Xiao Gao, Chenyang Wu, Mingjun Cao et al.

Decision Transformer (DT), which employs expressive sequence modeling techniques to perform action generation, has emerged as a promising approach to offline policy optimization. However, DT generates actions conditioned on a desired future return, which is known to bear some weaknesses such as the susceptibility to environmental stochasticity. To overcome DT's weaknesses, we propose to empower DT with dynamic programming. Our method comprises three steps. First, we employ in-sample value iteration to obtain approximated value functions, which involves dynamic programming over the MDP structure. Second, we evaluate action quality in context with estimated advantages. We introduce two types of advantage estimators, IAE and GAE, which are suitable for different tasks. Third, we train an Advantage-Conditioned Transformer (ACT) to generate actions conditioned on the estimated advantages. Finally, during testing, ACT generates actions conditioned on a desired advantage. Our evaluation results validate that, by leveraging the power of dynamic programming, ACT demonstrates effective trajectory stitching and robust action generation in spite of the environmental stochasticity, outperforming baseline methods across various benchmarks. Additionally, we conduct an in-depth analysis of ACT's various design choices through ablation studies. Our code is available at https://github.com/LAMDA-RL/ACT.

Chenyang Wu, Lina Lei, Fan Li et al.

Recent advances in Diffusion Transformer (DiT)-based video generation technologies have shown impressive results for video object removal. However, these methods still suffer from substantial inference latency. For instance, although MiniMax Remover achieves state-of-the-art visual quality, it operates at only around 10FPS, primarily due to dense computations over the entire spatiotemporal token space, even when only a small masked region actually requires processing. In this paper, we present YOSE, You Only Select Essential Tokens, an efficient fine-tuning framework. YOSE introduces two key components: Batch Variable-length Indexing (BVI) and Diffusion Process Simulator (DiffSim) Module. BVI is a differentiable dynamic indexing operator that adaptively selects essential tokens based on mask information, enabling variable-length token processing across samples. DiffSim provides a diffusion process approximation mechanism for unmasked tokens, which simulates the influence of unmasked regions within DiT self-attention to maintain semantic consistency for masked tokens. With these designs, YOSE achieves mask-aware acceleration, where the inference time scales approximately linearly with the masked regions, in contrast to full-token diffusion methods whose computation remains constant regardless of the mask size. Extensive experiments demonstrate that YOSE achieves up to 2.5X speedup in 70% of cases while maintaining visual quality comparable to the baseline. Code is available at: https://github.com/Wucy0519/YOSE-CVPR26.

Chenyang Wu, Jiayi Fu, Chun-Le Guo et al.

Due to large pixel movement and high computational cost, estimating the motion of high-resolution frames is challenging. Thus, most flow-based Video Frame Interpolation (VFI) methods first predict bidirectional flows at low resolution and then use high-magnification upsampling (e.g., bilinear) to obtain the high-resolution ones. However, this kind of upsampling strategy may cause blur or mosaic at the flows' edges. Additionally, the motion of fine pixels at high resolution cannot be adequately captured in motion estimation at low resolution, which leads to the misalignment of task-oriented flows. With such inaccurate flows, input frames are warped and combined pixel-by-pixel, resulting in ghosting and discontinuities in the interpolated frame. In this study, we propose a novel VFI pipeline, VTinker, which consists of two core components: guided flow upsampling (GFU) and Texture Mapping. After motion estimation at low resolution, GFU introduces input frames as guidance to alleviate the blurring details in bilinear upsampling flows, which makes flows' edges clearer. Subsequently, to avoid pixel-level ghosting and discontinuities, Texture Mapping generates an initial interpolated frame, referred to as the intermediate proxy. The proxy serves as a cue for selecting clear texture blocks from the input frames, which are then mapped onto the proxy to facilitate producing the final interpolated frame via a reconstruction module. Extensive experiments demonstrate that VTinker achieves state-of-the-art performance in VFI. Codes are available at: https://github.com/Wucy0519/VTinker.

Chenyang Wu, Yifan Duan, Xinran Zhang et al.

Localization and mapping are critical tasks for various applications such as autonomous vehicles and robotics. The challenges posed by outdoor environments present particular complexities due to their unbounded characteristics. In this work, we present MM-Gaussian, a LiDAR-camera multi-modal fusion system for localization and mapping in unbounded scenes. Our approach is inspired by the recently developed 3D Gaussians, which demonstrate remarkable capabilities in achieving high rendering quality and fast rendering speed. Specifically, our system fully utilizes the geometric structure information provided by solid-state LiDAR to address the problem of inaccurate depth encountered when relying solely on visual solutions in unbounded, outdoor scenarios. Additionally, we utilize 3D Gaussian point clouds, with the assistance of pixel-level gradient descent, to fully exploit the color information in photos, thereby achieving realistic rendering effects. To further bolster the robustness of our system, we designed a relocalization module, which assists in returning to the correct trajectory in the event of a localization failure. Experiments conducted in multiple scenarios demonstrate the effectiveness of our method.

Lei Song, Chenxiao Gao, Ke Xue et al.

Black-Box Optimization (BBO) has found successful applications in many fields of science and engineering. Recently, there has been a growing interest in meta-learning particular components of BBO algorithms to speed up optimization and get rid of tedious hand-crafted heuristics. As an extension, learning the entire algorithm from data requires the least labor from experts and can provide the most flexibility. In this paper, we propose RIBBO, a method to reinforce-learn a BBO algorithm from offline data in an end-to-end fashion. RIBBO employs expressive sequence models to learn the optimization histories produced by multiple behavior algorithms and tasks, leveraging the in-context learning ability of large models to extract task information and make decisions accordingly. Central to our method is to augment the optimization histories with \textit{regret-to-go} tokens, which are designed to represent the performance of an algorithm based on cumulative regret over the future part of the histories. The integration of regret-to-go tokens enables RIBBO to automatically generate sequences of query points that satisfy the user-desired regret, which is verified by its universally good empirical performance on diverse problems, including BBO benchmark functions, hyper-parameter optimization and robot control problems.

Chen-Xiao Gao, Chenyang Wu, Mingjun Cao et al.

Behavior regularization, which constrains the policy to stay close to some behavior policy, is widely used in offline reinforcement learning (RL) to manage the risk of hazardous exploitation of unseen actions. Nevertheless, existing literature on behavior-regularized RL primarily focuses on explicit policy parameterizations, such as Gaussian policies. Consequently, it remains unclear how to extend this framework to more advanced policy parameterizations, such as diffusion models. In this paper, we introduce BDPO, a principled behavior-regularized RL framework tailored for diffusion-based policies, thereby combining the expressive power of diffusion policies and the robustness provided by regularization. The key ingredient of our method is to calculate the Kullback-Leibler (KL) regularization analytically as the accumulated discrepancies in reverse-time transition kernels along the diffusion trajectory. By integrating the regularization, we develop an efficient two-time-scale actor-critic RL algorithm that produces the optimal policy while respecting the behavior constraint. Comprehensive evaluations conducted on synthetic 2D tasks and continuous control tasks from the D4RL benchmark validate its effectiveness and superior performance.

Zhe Xin, Chenyang Wu, Penghui Huang et al.

The recently developed Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) have shown encouraging and impressive results for visual SLAM. However, most representative methods require RGBD sensors and are only available for indoor environments. The robustness of reconstruction in large-scale outdoor scenarios remains unexplored. This paper introduces a large-scale 3DGS-based visual SLAM with stereo cameras, termed LSG-SLAM. The proposed LSG-SLAM employs a multi-modality strategy to estimate prior poses under large view changes. In tracking, we introduce feature-alignment warping constraints to alleviate the adverse effects of appearance similarity in rendering losses. For the scalability of large-scale scenarios, we introduce continuous Gaussian Splatting submaps to tackle unbounded scenes with limited memory. Loops are detected between GS submaps by place recognition and the relative pose between looped keyframes is optimized utilizing rendering and feature warping losses. After the global optimization of camera poses and Gaussian points, a structure refinement module enhances the reconstruction quality. With extensive evaluations on the EuRoc and KITTI datasets, LSG-SLAM achieves superior performance over existing Neural, 3DGS-based, and even traditional approaches. Project page: https://lsg-slam.github.io.

Bing Song, Jianing Liu, Sisi Jian et al.

Large language models (LLMs) have made significant strides, extending their applications to dialogue systems, automated content creation, and domain-specific advisory tasks. However, as their use grows, concerns have emerged regarding their reliability in simulating complex decision-making behavior, such as risky decision-making, where a single choice can lead to multiple outcomes. This study investigates the ability of LLMs to simulate risky decision-making scenarios. We compare model-generated decisions with actual human responses in a series of lottery-based tasks, using transportation stated preference survey data from participants in Sydney, Dhaka, Hong Kong, and Nanjing. Demographic inputs were provided to two LLMs -- ChatGPT 4o and ChatGPT o1-mini -- which were tasked with predicting individual choices. Risk preferences were analyzed using the Constant Relative Risk Aversion (CRRA) framework. Results show that both models exhibit more risk-averse behavior than human participants, with o1-mini aligning more closely with observed human decisions. Further analysis of multilingual data from Nanjing and Hong Kong indicates that model predictions in Chinese deviate more from actual responses compared to English, suggesting that prompt language may influence simulation performance. These findings highlight both the promise and the current limitations of LLMs in replicating human-like risk behavior, particularly in linguistic and cultural settings.