2D image-based 3D shape retrieval (2D-to-3D) aims at searching the corresponding 3D shapes (unlabeled) when given a 2D image (labeled), which is a fundamental task in computer vision and has gained a surge of attention in recent years. However, extensive prior works are limited by two settings, 1) reducing domain discrepancy while ignoring the 3D shape style, 2) 3D shapes are simply and brutally pseudo-annotated by the 2D image-supervised classifier, neglecting the structure information underlying the 3D shape domain. To remedy these issues, we propose a feature transformation framework with selective pseudo-labeling (FTSPL) for 2D-to-3D task. Specifically, we first employ CNNs to produce both 2D image and 3D shape (described as multiple views) features, then we force the inter-domain centroid alignment class-wisely to reduce the overall domain discrepancy. In addition to this, we further exploit the intra-category attribute variation (covariance) of 3D shape features to transform the 2D image features. By doing so, we can equip 2D features with 3D shape style. Since the centroid and covariance estimation of 3D shape features require accurate label predictions, we put forward a selective pseudo-labeling module, which can assign reliable pseudo-labels for 3D shapes via nearest category centroid and cluster analysis, respectively, while preserving the structure information of 3D shapes. Comprehensive experiments validate that our model surpasses the state-of-the-arts on standard 2D-to-3D benchmarks (MI3DOR and MI3DOR-2).

Given a 2D image query and a pool of 3D objects, the goal of image-object retrieval is to rank the 3D objects according to how well their content fits the query. Previous methods usually project 2D images and 3D objects into a joint embedding space and minimize the distance metric to complete the retrieval task. Since 2D images and 3D objects come from two different domains with large discrepancy, even when 3D objects and 2D images are mapped to a shared space, the gap in feature distribution remains significant, which always leads to domain misalignment. In this work, we propose a novel image-object retrieval method by leveraging optimal transport theory. Specifically, to tackle the dimensionality gap between 2D images and 3D objects, we first represent a 3D object via a sequence of its 2D projections. We then design a Cross-Domain View Attention module (CDVA) to automatically compute the optimal combination of 3D object projections given a 2D query image. Next, we exploit Weighted Optimal Transport (WOT)-based distance to depict the discrepancy between 2D images and 3D objects, and reduce the discrepancy to achieve instance-level alignment. Through this scheme, the transported 2D images and 3D objects with the same label are enforced to follow similar distributions. Finally, we design an explicit Category Centroid Alignment module (CCA) to achieve class-level alignment to improve the retrieval performance. Extensive experiments show that our method can achieve competitive performance on the MI3DOR and MI3DOR-2 benchmarks.

Legal Judgment Prediction aims to automatically infer judicial outcomes—including accusation charges, relevant statutes, and sentencing terms—from factual descriptions of criminal cases. Although large pretrained transformer models such as BERT achieve strong performance, their substantial computational cost and parameter scale hinder deployment in resource-constrained judicial environments. Existing lightweight architectures improve efficiency but often struggle to simultaneously capture global semantic context and fine-grained structural dependencies in lengthy legal texts.To address this challenge, we propose a Frequency-Guided Cross-View Interaction Network (FGCVIN), a lightweight framework that models complementary representations from multiple structural perspectives. FGCVIN first learns multi-scale semantic representations through a Global Contextual Abstraction Encoder and a Local Dependency Aggregation Encoder, which capture global contextual patterns and coherent local dependencies within legal fact descriptions. To introduce structural priors without increasing parameter complexity, a Frequency-Guided Prior Encoder extracts informative low-frequency signals via windowed Fast Fourier Transform and statistical pooling. Building upon these representations, a Cross-View Residual Attention Fusion module adaptively integrates semantic and structural features through gated residual fusion and attention refinement. A progressive three-stage training strategy further stabilizes the interaction among these components.Experiments on the CAIL benchmark datasets show that FGCVIN consistently outperforms several representative lightweight architectures demonstrating its effectiveness for resource-constrained legal intelligence systems.