In this work, we address the task of table image to LaTeX code generation, with the goal of automating the reconstruction of high-quality, publication-ready tables from visual inputs. A central challenge of this task lies in accurately handling complex tables—those with large sizes, deeply nested structures, and semantically rich or irregular cell content—where existing methods often fail. We begin with a comprehensive analysis, identifying key challenges and highlighting the limitations of current evaluation protocols. To overcome these issues, we propose a reinforced multimodal large language model (MLLM) framework, where a pre-trained MLLM is fine-tuned on a large-scale table-to-LaTeX dataset. To further improve generation quality, we introduce a dual-reward reinforcement learning strategy based on Group Relative Policy Optimization (GRPO). Unlike standard approaches that optimize purely over text outputs, our method incorporates both a structure-level reward on LaTeX code and a visual fidelity reward computed from rendered outputs, enabling direct optimization of the visual output quality. We adopt a hybrid evaluation protocol combining TEDS-Structure and CW-SSIM, and show that our method achieves state-of-the-art performance, particularly on structurally complex tables, demonstrating the effectiveness and robustness of our approach.

Anomaly detection in the hyperspectral image (HSI) has gradually become a hot topic in remote sensing. Recently, some tensor-based methods have been proposed to improve detection performance by exploiting the characteristic of HSI data existing in the inherent multi-dimensional. However, the existing tensor-based methods may only partially use the prior properties in both spatial and spectral dimensions. In this paper, we proposed a novel tensor ring (TR) decomposition with factors TV regularization model for hyperspectral anomaly detection. First, raw HSI data is decomposed into background and anomaly tensors. The tensor ring decomposition is adopted to exploit the low-rank property of the background existing in both spatial and spectral dimensions. Then, the total variation (TV) regularization is designed on the three dimensions of the background tensor to explore the piecewise smoothness of the background existing in both spatial and spectral dimensions. Further, this TV regularization is transferred to each factor tensor by exploiting the relationship between the background tensor and each factor tensor. Next, thel2,1norm regularization is designed on the anomaly tensor to exploit the group sparsity of anomaly pixels. Finally, the alternating direction method of multipliers (ADMM) scheme is adopted to update the involved variables. Experimental results validated on several real hyperspectral datasets demonstrate the effectiveness of the proposed algorithm.