CS Paper Radar

SE-MoLoRA: Shared-Expert LoRA Adapters for Domain-Specific Photographic Assessment

Bishwash Khanal, Anlan Zhang, Sasu Tarkoma, Tommi Mikkonen, Abhishek Kumar

cs.CV cs.MM ·submitted 2026-08-18 08:40 UTC · v2

Abstract

Vision-language models can describe images fluently, but they often fail to provide actionable photographic critique because semantic content and aesthetic judgment remain entangled. We propose SE-MoLoRA, a modular parameter-efficient adaptation framework for domain-specific photographic assessment. The method separates general photographic knowledge from specialist residual judgments using an always-active shared LoRA expert and routed adapters for composition, lighting, and technical quality. A lightweight query router selects the relevant specialist, enabling targeted critique without training separate full models. A rank-64 shared adapter captures broad photographic vocabulary, while rank-32 specialists learn domain-specific residuals with an orthogonal regularization penalty that encourages disentangled representations. Training data is obtained by distilling the Reddit Photo Critique Dataset into domain-labeled critique samples. On held-out critique generation, SE-MoLoRA improves BERTScore-F1 from 0.2317 to 0.4215 over monolithic LoRA and is preferred in 84.6\% of pairwise comparisons, while using fewer active parameters than separate specialist models. SVD-based ablation study shows that shared-specialist decomposition and orthogonal regularization reduce expert overlap. These results demonstrate that modular adaptation improves controllability and specificity in multimodal photographic critique.

Closest papers

Nearest neighbours by meaning across the whole index.

PortraitAes: Intent-Conditioned Structured Portrait Aesthetic Assessment

cs.CV ·2026-10-04 ·Junzhou Xie, Haozhong Xiong, Xunyun Tian, Kaile Du, Tianchen Yu +6 more
benchmark similarity 0.69
Abstract

Portrait aesthetic assessment assigns comparable scores according to how effectively human-centered images fulfill their photographic intent. These scores support data filtering, candidate selection, and preference modeling in image-generation pipelines. Existing methods typically predict a single aesthetic score or use general-purpose MLLMs without conditioning on photographic intent. This omission matters because the same blur, pose, lighting, or framing choice may serve one photographic intent but undermine another. These models thus learn context-agnostic aesthetic priors and yield inconsistent, inaccurate, misleading judgments for portraits with distinct photographic objectives. We introduce PortraitAes-Bench, an 11K-scale benchmark that decomposes this task into intent-conditioned subjudgments. Expert-authored rubrics define nine photographic intents, six first-level dimensions, and 22 secondary criteria. They support a structured pipeline for intent routing, specialist assessment, verification, and score fusion. Following this structure, we train PortraitAes with multi-task supervision. We then improve score comparability through Gaussian score calibration and within-dimension cross-image ranking. On the standard benchmark, PortraitAes achieves a Pearson correlation of 0.924 and a Spearman rank correlation of 0.934. On the hard-case set, its Pearson correlation is 0.829 and its Spearman rank correlation is 0.795. Across both sets, PortraitAes outperforms the evaluated general-purpose MLLMs and specialized aesthetic baselines.

0.22

AesCanvas: A Large-Scale Dataset and Benchmark for Aesthetic Critique and Contextual Suitability

cs.CV cs.AI ·2026-08-27 ·Xuanwei Hu, Haoyu Dong, Kejun Wu, Tianyi Liu, Jianjun Gao
benchmark similarity 0.66
Abstract · 10 pages, 4 figures, 6 tables. Supplementary material included

Recent advances in Multimodal Large Language Models (MLLMs) have extended Image Aesthetic Assessment (IAA) beyond scalar scores toward interpretable critique and guidance. Yet existing benchmarks mainly assess intrinsic visual quality or fixed domain criteria, leaving open whether an appealing image is appropriate for a specific purpose, audience, cultural setting, or domain convention. We introduce AesCanvas, a unified suite with two complementary components: CritiqueCanvas with 519,136 instruction-response pairs from 54,300 images supports long-form, multi-dimensional critique across photography, painting, and virtual imagery, whereas ContextCanvas with 301 expert-reviewed use scenarios evaluates contextual aesthetic suitability in realistic use scenarios. Under a unified protocol, we evaluate closed-source frontier, open-weight general, and aesthetic-specific MLLMs. Results reveal a clear separation between critique generation and context-sensitive judgment: reference-based lexical and semantic metrics only partially capture critique quality, while aesthetic specialists remain competitive on selected critique metrics yet substantially lag strong general-purpose MLLMs on ContextCanvas. Further analyses show that aesthetic specialization does not reliably transfer to contextual suitability and that model decisions may fail to track or ground themselves in decisive contextual visual cues. These findings establish culturally situated, evidence-grounded suitability as a distinct objective for aesthetic modeling.

0.33

LUTBench: Benchmarking Text-Driven Photographic Look Transfer

cs.CV ·2026-11-17 ·Chuanzhi Xu, Boyu Wei, Chengkun Yue, Xuanhua Yin, Yiyang Li +3 more
similarity 0.64
Abstract

Photographic color grading shapes the tone and color of a photograph to achieve a desired look. Text-driven photographic look transfer lets users specify these adjustments in natural language, with the goal of reproducing the requested look across photographs while preserving their content. However, prior research on text-driven color grading uses different datasets, text instructions, target definitions, and metrics. These differences make it difficult to compare how faithfully the methods reproduce the same look. To compare methods under a shared protocol, we introduce LUTBench. We collect 91 three-dimensional lookup tables (3D LUTs) and define one photographic look per LUT with a specified input conversion. Applying these looks to 5,000 MIT-Adobe FiveK photographs produces 455,000 input-target image pairs with reproducible target images. For each look, we use templates to turn measured tone and color changes on fixed color samples into candidate clauses. We use training image measurements to verify applicable clauses before composing the text description. The evaluation measures target fidelity, tests sensitivity to predefined input conversions, and separates generalization to unseen looks from reproduction of seen looks on new photographs. Across 16 methods, stronger text alignment does not guarantee target fidelity. However, most supervised variants improve fidelity on both seen and unseen looks.

0.32

KnowVis: Dual-View Evaluation and Diagnosis of World-Knowledge Grounding in Text-to-Image Models

cs.CV ·2026-11-13 ·Yonghang Yu, Ruoyu Wang, Haoyu Zhang, Lin Qi, Yawei Zhou +1 more
top venuebenchmarkcode/data similarity 0.63
Abstract · Accepted to NeurIPS 2026 E&D Track. Codebase: https://github.com/QuantumWhisper42/KnowVis

Text-to-image (T2I) models have made substantial progress in visual realism, aesthetic quality, and instruction following. However, real-world prompts often go beyond explicit visual descriptions and require implicit facts, structured knowledge, and domain-specific commonsense. Existing evaluations mainly focus on explicit prompt-to-image semantic alignment or specific knowledge subdomains, lacking unified evaluation and diagnosis of world-knowledge grounding in T2I models. We introduce KnowVis for dual-view evaluation and diagnosis of world-knowledge grounding in T2I generation. Built on a hierarchical knowledge taxonomy with case-level annotations, KnowVis combines two complementary benchmark views with a structured MLLM-based VQA judge protocol. The Skill-Tree Benchmark supports fine-grained atomic knowledge evaluation, while the Generalized Benchmark examines open-ended knowledge composition and expression. The protocol uses Required VQAs to assess atomic knowledge correctness in the Skill-Tree Benchmark, and combines Required, Optional, and dynamically discovered knowledge expressions with human-judge collaborative evaluation in the Generalized Benchmark. We validate the protocol's reliability on the Skill-Tree validation split, enabling scalable evaluation and diagnostic error attribution. Experiments show that KnowVis reveals notable gaps in world-knowledge grounding across existing T2I models and provides structured diagnostic signals for knowledge-intensive image generation.

0.52

Harnessing Domain Specialists in Multimodal Mixture-of-Experts for Efficient Adaptation

cs.CV ·2026-10-01 ·Damiano Marsili, Raphi Kang, Aditya Mehta, Pietro Perona, Georgia Gkioxari
code/data similarity 0.63
Abstract · Project page: https://glab-caltech.github.io/expertlens/

Mixture-of-Experts (MoE) architectures scale model capacity through sparse computation, routing each token through only a small subset of experts. In this work, we explore whether this sparsity gives rise to emergent intrinsic organization in multimodal MoEs. We find that experts develop strong semantic specialization across modalities and domains despite not being explicitly trained for modularity. Building on this structure, we introduce ExpertLens, a data-free method that identifies domain-specialized experts directly from pretrained model weights by decoding router weights into semantically meaningful vocabulary tokens. We leverage this specialization for efficient multimodal adaptation by selectively fine-tuning experts relevant to a target domain. Across math, medical, and remote sensing tasks, ExpertLens matches or surpasses full fine-tuning while updating only 21.7 - 47.0% of model parameters and achieving a 4.0x average training speedup, and outperforms LoRA in both adaptation performance and training efficiency. These results show that sparsity introduced for efficiency can give rise to semantic modularity that is directly useful for efficient adaptation.

0.29

Beyond Domain-Level Adaptation: Margin-Oriented Semantic-Appearance Interaction Correction for Personalized Federated Vision-Language Models

cs.CV ·2026-10-01 ·Wentao Yue, Qingyu Mao, Tianyou Lai, Ahmed M. Abdelmoniem, Qilei Li
similarity 0.63
Abstract

Federated parameter-efficient fine-tuning enables distributed clients to adapt pretrained vision-language models without sharing raw data or updating the full backbone. Its effectiveness, however, is limited by domain heterogeneity across clients. Existing personalized methods separate globally shared knowledge from client-specific style, but they largely treat each domain as a class-agnostic transformation. We show that this abstraction is insufficient: the cross-domain displacement associated with a fixed domain varies across semantic classes, and only a subset of these class-domain residuals damages the image-text decision margin. We therefore propose Margin-Oriented Semantic-Appearance Interaction Correction (MOSAIC), which first constructs a decision-aware harmfulness score that measures whether a training-derived class-domain residual favors a competing text prototype over the true class. It then models fine-grained class-domain interactions with a low-rank residual adapter whose class factors and residual basis are globally shared while domain factors remain client-private. An image-conditioned gate further controls candidate-wise correction, and harmful-pair-aware reweighting prioritizes decision-relevant residuals during local optimization. Extensive experiments on Office31, OfficeHome, and DomainNet100 demonstrate that MOSAIC consistently improves macro-client top-1 accuracy across all evaluated domain-shift and joint domain-label-shift settings.

0.23

ProSE: Probing Reconstruction Error for Semantic-guided Expert Skipping in MoE-MLLMs

cs.CV ·2026-11-16 ·Jiameng Li, Minye Wu, Matthew B. Blaschko
similarity 0.62
Abstract · 8 pages

Mixture-of-Experts Multimodal Large Language Models (MoE-MLLMs) offer scalable inference through sparse expert activation, yet substantial computational redundancy remains among the activated experts. Existing expert skipping methods typically calibrate thresholds based on target skipping ratios or final-output divergence, overlooking the accumulation of skipping-induced error across layers and the varying semantic importance of individual tokens. To address that, we propose ProSE, a training-free, semantic-guided expert skipping framework that probes layer-wise reconstruction error to promote representation consistency compared with default activation. Specifically, we construct query-aware semantic masks to identify salient visual tokens and progressively calibrate modality-specific skipping thresholds for each MoE layer. By minimizing masked reconstruction error under a given budget, ProSE maintains semantic consistency after skipping and allows adaptive expert allocation under limited activation budgets. Extensive experiments on 9 benchmarks demonstrate consistent improvements on average over existing expert skipping methods, particularly under aggressive skipping settings. ProSE maintains over 96\% performance on KimiVL-A3B-Inst with 83\% expert skipping and accelerates prefilling by $1.7\times$.

0.19

Towards Efficient Multimodal and Multilingual Opinion Extraction for STI: A QLoRA-Based Fine-Tuning Approach

cs.AI ·2026-08-14 ·Sheng Hong, Xuanqi Wang, Jiacheng Wang, Yuwei Wang
similarity 0.62
Abstract

Recent advances in large language models (LLMs) have reshaped semantic analysis. Opinion Extraction (OE) for Science and Technology Intelligence (STI) requires concise core opinions from large information streams. Off-the-shelf models struggle to filter noise from these streams and show limited structured-output reliability in zero-shot multilingual and multi-modal settings. To address information overload and extraction defocus, this study proposes a multimodal core-opinion extraction framework in which visual evidence serves as a contextual anchor for textual judgment. Using VideoLLaMA2 (VL2) and VideoLLaMA2.1 (VL2.1) as the base models, we apply Quantized Low-Rank Adaptation (QLoRA) fine-tuning on a curated dataset of 2,194 multilingual and multimodal samples. Under the selected Image-Augmented setting, fine-tuned VL2.1 generates structured JSON core-opinion outputs, achieving 64.98% Precision, 42.15% Recall, 51.14% F1-score, and 74.00% sample-level accuracy. Relative to the zero-shot VL2.1 setting, it raises the F1-scores of Spanish and Russian from 4.83% and 0.45% to 46.05% and 51.93%, respectively. The framework further incorporates a Fuzzy Cumulative Prospect Theory-based post-extraction triage module for case-level value assessment, providing a case-level value signal for downstream STI screening.

0.04