Meng Wang

This work aims to tackle the all-in-one image restoration task, which seeks to handle multiple types of degradation with a single model. The primary challenge is to extract degradation representations from the input degraded images and use them to guide the model's adaptation to specific degradation types. Building on the insight that various degradations affect image content differently across frequency bands, we propose a new dual-transformer approach comprising two components: a frequency-aware degradation estimation transformer (Dformer) and a degradation-adaptive restoration transformer (Rformer). The Dformer captures the essential characteristics of various degradations by decomposing the input into different frequency components. By understanding how degradations affect these frequency components, the Dformer learns robust priors that effectively guide the restoration process. The Rformer then employs a degradation-adaptive self-attention (DA-SA) module to selectively focus on the most affected frequency components, guided by the learned degradation representations. Extensive experimental results demonstrate that our approach outperforms existing methods in five representative restoration tasks, including denoising, deraining, dehazing, deblurring, and low-light enhancement. In addition, our method offers benefits for handling real-world degradations, spatially variant degradations, and unseen degradation levels.

Recent adaptations can boost the low-shot capability of Contrastive Language-Image Pre-training (CLIP) by effectively facilitating knowledge transfer. However, these adaptation methods are usually operated on the global view of an input image, and thus biased perception of partial local details of the image. To solve this problem, we propose a Visual Content Refinement (VCR) before the adaptation calculation during the test stage. Specifically, we first decompose the test image into different scales to shift the feature extractor’s attention to the details of the image. Then, we select the image view with the max prediction margin in each scale to filter out the noisy image views, where the prediction margins are calculated from the pre-trained CLIP model. Finally, we merge the content of the aforementioned selected image views based on their scales to construct a new robust representation. Thus, the merged content can be directly used to help the adapter focus on both global and local parts without any extra training parameters. We apply our method to 3 popular low-shot benchmark tasks with 13 datasets and achieve a significant improvement over state-of-the-art methods. For example, compared to the baseline (Tip-Adapter) on the few-shot classification task, our method achieves about 2% average improvement for both training-free and training-needed settings. Our code is available at: https://github.com/injadlu/VCR .

User-generated reviews contain rich semantics that can reveal users’ fine-grained preferences beyond what interaction data alone can capture. However, existing review-based recommender systems often fail to achieve semantic alignment between review content and user–item interactions, as they either treat reviews as coarse textual signals or rely on sentiment-oriented heuristics. Achieving such fine-grained alignment is challenging due to the inherent complexity and context dependence of review texts, as well as the semantic gap between explicit review information and implicit interaction preferences. To bridge this gap, we propose LLM-ASAR , a L arge L anguage M odel-driven framework for A spect-based S emantic A lignment in Review-based R ecommendation that explicitly aligns aspect-level review semantics with user–item interaction patterns. Specifically, LLM-ASAR leverages LLMs with Chain-of-Thought (CoT) prompting to perform sentence-level reasoning on user reviews, extracting aspect-specific semantics that are structurally integrated into aspect-specific interaction graphs. A multi-aspect contrastive alignment mechanism is further introduced to align user and item embeddings with corresponding aspect-level review features, ensuring consistency across modalities. Extensive experiments on three real-world datasets demonstrate that LLM-ASAR not only achieves significant accuracy improvements over state-of-the-art baselines but also yields interpretable insights into user preferences through its aspect-aware design. Our data and code are available at https://github.com/HuilinChenJN/LLM-ASAR.