Feature Fusion

Follow this topic Ask about this topic

Momentum

12 papers in the last four weeks, against 2 the four weeks before. 0.1% of all new papers.

Jul 13Week of Sep 28

Latest papers 77

May 11, 2026cs.CV

DetRefiner: Model-Agnostic Detection Refinement with Feature Fusion Transformer

Open-vocabulary object detection (OVOD) aims to detect both seen and unseen categories, yet existing methods often struggle to generalize to novel objects due to limited integration of global and local contextual cues. We propose DetRefiner, a simple yet effective plug-and-play framework that learns to fuse global and local features to refine open-vocabulary detection. DetRefiner processes global image features and patch-level image features from foundational models (e.g., DINOv3) through a lightweight Transformer encoder. The encoder produces a class vector capturing image-level attributes and patch vectors representing local region attributes, from which attribute reliability is inferred to recalibrate the base model's confidence. Notably, DetRefiner is trained independently of the base OVOD model, requiring neither access to its internal features nor retraining. At inference, it operates solely on the base detector's predictions, producing auxiliary calibration scores that are merged with the base detector's scores to yield the final refined confidence. Despite this simplicity, DetRefiner consistently enhances multiple OVOD models across COCO, LVIS, ODinW13, and Pascal VOC, achieving gains of up to +10.1 AP on novel categories. These results highlight that learning to fuse global and local representations offers a powerful and general mechanism for advancing open-world object detection. Our codes and models are available at https://github.com/hitachi-rd-cv/detrefiner.

Soichiro Okazaki, Tatsuya Sasaki, Hiroki Ohashi

Hitachi, Ltd. Research and Development Group, Japan

Open-Vocabulary Object Detection Object Detection Feature Fusion

May 9, 2026cs.LG

MicroFuse: Protein-to-Genome Expert Fusion for Microbial Operon Reasoning

Predicting microbial operon co-membership requires integrating two complementary biological signals: protein-scale molecular identity and genome-context organization. While recent biological foundation models provide powerful representations of each view independently, naive concatenation of these modalities ignores a key biological property -- protein identity and genomic context may agree when adjacent genes form a coherent functional module, or conflict when sequence similarity is misleading but genomic layout indicates independent regulation. We present MicroFuse, a protein-to-genome expert fusion framework that integrates structure-aware protein representations from ProstT5 with genome-context representations from Bacformer through a four-expert Mixture-of-Experts module (protein, genome-context, agreement, and conflict experts) with a learned soft router. Training combines binary cross-entropy with symmetric cross-modal InfoNCE alignment and disagreement-weighted supervised contrastive shaping. We further construct OG-Operon100K, a 100,000-pair scaffold-level benchmark from the OMG metagenomic corpus with biologically grounded positive and negative criteria. On OG-Operon100K, MicroFuse achieves the strongest AUROC, AUPRC, mAP, and mAR among ProstT5-only, Bacformer-only, and Concat MLP baselines. Ablations identify cross-modal contrastive alignment as the dominant component, and a hard sequence-conflict subset reveals MicroFuse's largest gains precisely in biologically ambiguous cases where protein identity alone is misleading.

Seungik Cho

Department of Physics and Astronomy, Rice University, Texas, USA.

Genomics Multimodal Fusion Feature Fusion

May 5, 2026cs.CV

Unified Multimodal Visual Tracking with Dual Mixture-of-Experts

Multimodal visual object tracking can be divided into to several kinds of tasks (e.g. RGB and RGB+X tracking), based on the input modality. Existing methods often train separate models for each modality or rely on pretrained models to adapt to new modalities, which limits efficiency, scalability, and usability. Thus, we introduce OneTrackerV2, a unified multi-modal tracking framework that enables end-to-end training for any modality. We propose Meta Merger to embed multi-modal information into a unified space, allowing flexible modality fusion and robustness. We further introduce Dual Mixture-of-Experts (DMoE): T-MoE models spatio-temporal relations for tracking, while M-MoE embeds multi-modal knowledge, disentangling cross-modal dependencies and reducing feature conflicts. With a shared architecture, unified parameters, and a single end-to-end training, OneTrackerV2 achieves state-of-the-art performance across five RGB and RGB+X tracking tasks and 12 benchmarks, while maintaining high inference efficiency. Notably, even after model compression, OneTrackerV2 retains strong performance. Moreover, OneTrackerV2 demonstrates remarkable robustness under modality-missing scenarios.

Lingyi Hong, Jinglun Li, Xinyu Zhou +6

Shanghai Key Lab of Intelligent Information Processing, College of Computer Science and Artificial Intelligence, Fudan University

Cross-Modal Learning Multimodal Robustness Unified Multimodal Models Mixture of Experts Visual Object Tracking Multimodal Fusion+1

May 4, 2026cs.CV

Unified Map Prior Encoder for Mapping and Planning

Online mapping and end-to-end (E2E) planning in autonomous driving remain largely sensor-centric, leaving rich map priors, including HD/SD vector maps, rasterized SD maps, and satellite imagery, underused because of heterogeneity, pose drift, and inconsistent availability at test time. We present UMPE, a Unified Map Prior Encoder that can ingest any subset of four priors and fuse them with BEV features for both mapping and planning. UMPE has two branches. The vector encoder pre-aligns HD/SD polylines with a frame-wise SE(2) correction, encodes points via multi-frequency sinusoidal features, and produces polyline tokens with confidence scores. BEV queries then apply cross-attention with confidence bias, followed by normalized channel-wise gating to avoid length imbalance and softly down-weight uncertain sources. The raster encoder shares a ResNet-18 backbone conditioned by FiLM with scaling and shift at every stage, performs SE(2) micro-alignment, and injects priors through zero-initialized residual fusion, so the network starts from a do-no-harm baseline and learns to add only useful prior evidence. A vector-then-raster fusion order reflects the inductive bias of geometry first, appearance second. On nuScenes mapping, UMPE lifts MapTRv2 from 61.5 to 67.4 mAP (+5.9) and MapQR from 66.4 to 71.7 mAP (+5.3). On Argoverse2, UMPE adds +4.1 mAP over strong baselines. UMPE is compositional: when trained with all priors, it outperforms single-prior models even when only one prior is available at test time, demonstrating powerset robustness. For E2E planning with the VAD backbone on nuScenes, UMPE reduces trajectory error from 0.72 to 0.42 m L2 on average (-0.30 m) and collision rate from 0.22% to 0.12% (-0.10%), surpassing recent prior-injection methods. These results show that a unified, alignment-aware treatment of heterogeneous map priors yields better mapping and better planning.

Zongzheng Zhang, Sizhe Zou, Guantian Zheng +12

Institute for AI Industry Research (AIR), Tsinghua University. · Bosch Corporate Research, China. · Shanghai Jiao Tong University. +1

End-to-End Autonomous Driving Autonomous Driving Autonomous Driving Planning Feature Fusion

May 4, 2026cs.AI

Triple Spectral Fusion for Sensor-based Human Activity Recognition

The field of sensor-based human activity recognition (HAR) mainly uses posture, motion and context data of Inertial Measurement Units (IMUs) to identify daily activities. Despite the advancements in learning-based methods, it is challenging to perform information fusion from the temporal perspective due to the complexities in fusing heterogeneous sensor data and establishing long-term context correlations. This paper proposes a novel triple spectral fusion framework tailored for HAR. First, we develop an adaptive complementary filtering technique for noise suppression and organize each IMU's sensors into posture and motion modality nodes. Given that IMU nodes form a dynamic heterogeneous graph, we then apply adaptive filtering within the graph Fourier domain to merge both homogeneous and heterogeneous node information. Furthermore, an adaptive wavelet frequency selection approach is implemented to suppress context redundancy and shorten the length of features. This approach enhances both timestamp-based graph aggregation and the correlation of long-term contexts. Our framework uses adaptive filtering in the Fourier, graph Fourier, and wavelet domains, enabling effective multi-sensor fusion and context correlation. Extensive experiments on ten benchmark datasets demonstrate the superior performance of our framework. Project page: https://github.com/crocodilegogogo/TSF-TPAMI2026.

Ye Zhang, Longguang Wang, Qing Gao +3

School of Electronics and Communication Engineering, the Shenzhen Campus of Sun Yat-sen University, Sun Yat-sen University, Shenzhen 518107, China · Aviation University of Air Force, Changchun 130012, China · College of Computer Science, Chongqing University, Chongqing 400044, China +1

Graph Signal Processing Human Activity Recognition Frequency-Domain Feature Learning Multimodal Graph Learning Feature Fusion Multimodal Sensor Fusion

May 4, 2026cs.CV

M\textsuperscript{4}Fuse: Lightweight State-Space MoE with a Cross-Scale Gating Bridge for Brain Tumor Segmentation

Encoder-decoder imbalance and the reliance on large input volumes make many 3D brain tumor segmentation models both compute-heavy and brittle. We present M\textsuperscript{4}Fuse, a lightweight network that prioritizes discriminative brain tumor cues over exhaustive appearance reconstruction. Our method balances encoder and decoder capacity and replaces depth expansion with a synergistic design: it propagates long-range context with linear complexity via a grouped state space mixer, denoises and aligns skip features using a cross-scale dual-stage gating bridge, and absorbs cross-site acquisition shifts with a sample-level mixture-of-experts. On the BraTS2019 and BraTS2021 benchmarks, M\textsuperscript{4}Fuse outperforms other lightweight excellent methods in both parameter count and performance. Even at a challenging input resolution of

64\times128\times128

(half that of existing excellent models), M\textsuperscript{4}Fuse reduces parameters by 62.63% and improves average performance by 0.09%. Ablations of key components validate the method's exceptional parameter-to-accuracy efficiency and robustness across diverse data centers.

Meihua Zhou, Xinyu Tong, Li Yang

University of Chinese Academy of Sciences · 2Wannan Medical University

Image Segmentation Multi-Scale Feature Fusion Medical Image Segmentation 3D Medical Image Segmentation Feature Fusion Brain Tumor Segmentation+1

May 3, 2026cs.LG

MER-DG: Modality-Entropy Regularization for Multimodal Domain Generalization

Deploying multimodal models in real-world scenarios requires generalization to new environments where recording conditions differ from training, a challenge known as multimodal domain generalization (MMDG). Standard architectures employ separate encoders for each modality and a fusion module, training the system end-to-end by optimizing on the fused features. In this paper, we identify that such joint optimization causes encoders to exploit cross-modal co-occurrences, statistical relationships between modalities that arise from source-specific recording conditions, rather than learning domain-invariant features. We term this failure mode Fusion Overfitting. To address this, we propose Modality-Entropy Regularization for Domain Generalization (MER-DG), which maximizes the entropy of each encoder's feature distribution to preserve feature diversity. MER-DG is architecture-agnostic and integrates into existing multimodal frameworks as an additive loss term. Extensive experiments on EPIC-Kitchens and HAC benchmarks demonstrate average improvements of approximately 5% over standard fusion and approximately 2% over state-of-the-art methods.

Yavuz Yarici, Ghassan AlRegib

1OLIVES at the Center for Signal and Information Processing CSIP, School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, USA.

Domain Generalization Cross-Domain Generalization Multimodal Fusion Entropy Regularization Feature Fusion

May 1, 2026cs.CV

BOLT: Online Lightweight Adaptation for Preparation-Free Heterogeneous Cooperative Perception

Most existing heterogeneous cooperative perception methods depend on prior preparation like offline joint training or tailored collaborator-model adaptation. Such preprocessing is, however, generally impractical in real scenarios, as agents are usually independently trained by different developers and meet occasionally online. This work investigates \emph{preparation-free heterogeneous cooperative perception}, where agents use independently trained single-agent detectors without any pre-deployment coordination. We find direct cross-agent fusion under this setting greatly underperforms ego-only perception. We present BOLT, a lightweight plug-and-play module that adapts neighboring features online via ego-as-teacher distillation, requiring only ego predictions without ground-truth labels. BOLT leverages high-confidence ego perception features to guide cross-agent feature-domain alignment, while enabling neighbors to contribute features in the ego's low-confidence regions. With only 0.9M trainable parameters, BOLT improves AP@50 by up to 32.3 points over vanilla unadapted fusion in the preparation-free setting. It consistently outperforms ego-only results on DAIR-V2X and OPV2V, across different encoder pairs and fusion strategies. Code: https://github.com/sidiangongyuan/BOLT.

Kang Yang, Tianci Bu, Peng Wang +2

Renmin University of China · National University of Defense Technology

Cooperative Perception Adaptive Learning Heterogeneous Collaborative Perception Feature Fusion

Apr 30, 2026cs.CV

Linguistically Informed Multimodal Fusion for Vietnamese Scene-Text Image Captioning: Dataset, Graph Framework, and Phonological Attention

Scene-text image captioning requires fusing three information streams -- visual features, OCR-detected text, and linguistic knowledge -- to generate descriptions that faithfully integrate text visible in images. Existing fusion approaches treat text as language-agnostic, which fails for Vietnamese: a tonal language where diacritics alter word meaning, OCR errors are pervasive, and word boundaries are ambiguous. We argue that Vietnamese scene-text captioning demands \textit{linguistically informed multimodal fusion}, where language-specific structural knowledge is explicitly incorporated into the fusion mechanism. Motivated from these insights, we propose \textbf{HSTFG} (Heterogeneous Scene-Text Fusion Graph), a general-purpose graph fusion framework with learned spatial attention bias, and show through topology analysis that cross-modal graph edges are harmful for scene-text fusion. Building on this finding, we design \textbf{PhonoSTFG} (Phonological Scene-Text Fusion Graph) which specializes graph-level fusion for Vietnamese linguistic reasoning. To support evaluation, we introduce \textbf{ViTextCaps}, the first large-scale Vietnamese scene-text captioning dataset (\textbf{15{,}729} images with \textbf{74{,}970} captions), with comprehensive linguistic analysis showing that 52.8% of the vocabulary is at risk of diacritic collision.

Nhi Ngoc-Yen Nguyen, Anh-Duc Nguyen, Nghia Hieu Nguyen +2

Faculty of Information Science and Engineering · University of Information Technology · Vietnam National University, Ho Chi Minh, Viet Nam

Image Captioning Low-Resource Language Processing Multimodal Fusion Feature Fusion

Apr 28, 2026cs.CV

Control Your Queries: Heterogeneous Query Interaction for Camera-Radar Fusion

In autonomous driving, camera-radar fusion offers complementary sensing and low deployment cost. Existing methods perform fusion through input mixing, feature map mixing, or query-based feature sampling. We propose a new fusion paradigm, termed heterogeneous query interaction, and present ConFusion, a camera-radar 3D object detector. ConFusion combines image queries, radar queries, and learnable world queries distributed in 3D space to improve query initialization and object coverage. To encourage cross-type interaction among heterogeneous queries, we introduce heterogeneous query mixing (QMix), which performs dedicated cross-type attention after feature sampling to consolidate complementary object evidence. We further propose interactive query swap sampling (QSwap), which improves feature sampling by allowing related queries to exchange informative feature tokens under attention and geometric constraints. Experiments on the nuScenes dataset show that ConFusion achieves state-of-the-art performance, reaching 59.1 mAP and 65.6 NDS on the validation set, and 61.6 mAP and 67.9 NDS on the test set.

Jialong Wu, Yihan Wang, Matthias Rottmann

1Osnabrück University · 3Aptiv Services Deutschland GmbH · University of Wuppertal

Camera-Radar Fusion Autonomous Driving Perception 3D Object Detection Multimodal Fusion Feature Fusion

Apr 25, 2026cs.CV

A Heterogeneous Two-Stream Framework for Video Action Recognition with Comparative Fusion Analysis

Most two-stream action recognition networks apply the same convolutional backbone to both RGB and optical flow streams, ignoring the fact that the two modalities have fundamentally different structural properties. Optical flow captures fine-grained motion patterns, while RGB frames carry rich appearance and scene context - treating them identically discards this distinction. We propose DualStreamHybrid, a heterogeneous two-stream architecture that assigns each stream a backbone suited to its input: a pretrained ViT-Tiny/16 for RGB frames, and a MobileNetV2 trained from scratch on a 20-channel stacked optical flow representation. A learned projection layer maps the two differently-sized feature vectors to a common dimensionality before fusion, enabling the two streams to interact without forcing architectural symmetry. We design five fusion strategies within a unified framework - late fusion, concatenation, cross-attention, weighted fusion, and gated fusion - and evaluate them on UCF11 (1,600 videos, 11 classes) and UCF50 (6,681 videos, 50 classes) to study how fusion behaviour scales with dataset size. On UCF11, cross-attention achieves 98.12% test accuracy, outperforming the RGB-only ViT-Tiny baseline of 95.94%, which suggests that explicit inter-modal attention is particularly effective on smaller, less complex datasets. On UCF50, weighted fusion reaches 96.86% and proves the most consistent strategy across both benchmarks. The learned stream weights reveal an interesting pattern: UCF11 sees near-equal modality contribution (RGB: 0.507, flow: 0.493), while UCF50 favours the RGB stream slightly more (RGB: 0.554, flow: 0.446) - arguably reflecting the larger and more visually diverse action space. Taken together, these results suggest that even a lightweight motion stream meaningfully complements a strong appearance encoder, and that the optimal fusion strategy depends on dataset scale.

Md. Afzalur Rahaman, Tahmid Rahman

Department of Computer Science & Engineering Hamdard University Bangladesh Munshiganj, Bangladesh

Video Action Recognition Cross-Modal Attention Multimodal Fusion Feature Fusion

Apr 17, 2026cs.CV

Multilevel neural networks with dual-stage feature fusion for human activity recognition

Human activity recognition (HAR) refers to the process of identifying human actions and activities using data collected from sensors. Neural networks, such as convolutional neural networks (CNNs), long short-term memory (LSTM) networks, convolutional LSTM, and their hybrid combinations, have demonstrated exceptional performance in various research domains. Developing a multilevel individual or hybrid model for HAR involves strategically integrating multiple networks to capitalize on their complementary strengths. The structural arrangement of these components is a critical factor influencing the overall performance. This study explores a novel framework of a two-level network architecture with dual-stage feature fusion: late fusion, which combines the outputs from the first network level, and intermediate fusion, which integrates the features from both the first and second levels. We evaluated

15

different network architectures of CNNs, LSTMs, and convolutional LSTMs, incorporating late fusion with and without intermediate fusion, to identify the optimal configuration. Experimental evaluation on two public benchmark datasets demonstrates that architectures incorporating both late and intermediate fusion achieve higher accuracy than those relying on late fusion alone. Moreover, the optimal configuration outperforms baseline models, thereby validating its effectiveness for HAR.

Abeer FathAllah Brery, Ascensión Gallardo-Antolín, Israel Gonzalez-Carrasco +1

Human Activity Recognition Recurrent Neural Networks Convolutional LSTM Convolutional Neural Networks Feature Fusion

Apr 13, 2026cs.CV

Towards Automated Solar Panel Integrity: Hybrid Deep Feature Extraction for Advanced Surface Defect Identification

To ensure energy efficiency and reliable operations, it is essential to monitor solar panels in generation plants to detect defects. It is quite labor-intensive, time consuming and costly to manually monitor large-scale solar plants and those installed in remote areas. Manual inspection may also be susceptible to human errors. Consequently, it is necessary to create an automated, intelligent defect-detection system, that ensures continuous monitoring, early fault detection, and maximum power generation. We proposed a novel hybrid method for defect detection in SOLAR plates by combining both handcrafted and deep learning features. Local Binary Pattern (LBP), Histogram of Gradients (HoG) and Gabor Filters were used for the extraction of handcrafted features. Deep features extracted by leveraging the use of DenseNet-169. Both handcrafted and deep features were concatenated and then fed to three distinct types of classifiers, including Support Vector Machines (SVM), Extreme Gradient Boost (XGBoost) and Light Gradient-Boosting Machine (LGBM). Experimental results evaluated on the augmented dataset show the superior performance, especially DenseNet-169 + Gabor (SVM), had the highest scores with 99.17% accuracy which was higher than all the other systems. In general, the proposed hybrid framework offers better defect-detection accuracy, resistance, and flexibility that has a solid basis on the real-life use of the automated PV panels monitoring system.

Muhammad Junaid Asif, Muhammad Saad Rafaqat, Usman Nazakat +2

Artificial Intelligence Technology Centre (AITeC), National Centre for Physics (NCP), Islamabad 44000, Pakistan · School of Computing, Pak-Austria Fachhochschule: Institute of Applied Sciences and Technology (PAFIAST), Haripur 22620, Pakistan

Industrial Inspection Feature Fusion

Apr 2, 2026cs.CV

COMPASS: Fusion-Matched Supervision for Missing-Modality Human Sensing

Multimodal human activity recognition (HAR) and human pose estimation (HPE) must cope with modalities missing at inference. Completions generated for a fixed fusion model must preserve the fusion readout: what fusion uses from each modality. We propose COMPASS, a completion-and-fusion framework in which each modality occupies a fixed slot filled by an observed representation or a completion inferred from available inputs. Its core principle, fusion-matched supervision, supervises completions with same-sample real targets at readout granularity: token means for HAR, where fusion averages tokens, and per-joint representations for HPE, where fusion preserves joint structure. Trained with complete multimodal samples, COMPASS improves subset-averaged performance over the strongest baselines on XRF55 and MM-Fi for HAR and on MM-Fi for HPE. At fixed architecture, fusion-matched supervision outperforms full-token and centered-token matching in HAR, and joint-mean matching and no matching in HPE. The code is available at: https://github.com/haowangcoder/COMPASS.

Hao Wang, Yanyu Qian, Pengcheng Weng +4

Universität Bern, Switzerland · Xi’an Jiaotong University, China · Nanyang Technological University, Singapore

Missing-Modality Learning Human Activity Recognition Human Pose Estimation Multimodal Fusion Feature Fusion

Feb 9, 2026cs.RO

GaussianCaR: Gaussian Splatting for Efficient Camera-Radar Fusion

Robust and accurate perception of dynamic objects and map elements is crucial for autonomous vehicles performing safe navigation in complex traffic scenarios. While vision-only methods have become the de facto standard due to their technical advances, they can benefit from effective and cost-efficient fusion with radar measurements. In this work, we advance fusion methods by repurposing Gaussian Splatting as an efficient universal view transformer that bridges the view disparity gap, mapping both image pixels and radar points into a common Bird's-Eye View (BEV) representation. Our main contribution is GaussianCaR, an end-to-end network for BEV segmentation that, unlike prior BEV fusion methods, leverages Gaussian Splatting to map raw sensor information into latent features for efficient camera-radar fusion. Our architecture combines multi-scale fusion with a transformer decoder to efficiently extract BEV features. Experimental results demonstrate that our approach achieves performance on par with, or even surpassing, the state of the art on BEV segmentation tasks (57.3%, 82.9%, and 50.1% IoU for vehicles, roads, and lane dividers) on the nuScenes dataset, while maintaining a 3.2x faster inference runtime. Code and project page are available online.

Santiago Montiel-Marín, Miguel Antunes-García, Fabio Sánchez-García +3

Department of Electronics. University of Alcalá, Spain. · Department of Cognitive Robotics. Delft University of Technology, The Netherlands.

Camera-Radar Fusion Autonomous Driving Perception Gaussian Splatting Semantic Segmentation Feature Fusion

Nov 30, 2025cs.LG

FDRMFL: Multimodal Federated Feature Extraction Model Based on Information Maximization and Contrastive Learning

We propose FDRMFL, a task-driven multimodal feature extraction framework for federated regression under non-IID data distributions. Extracting predictive features from high-dimensional multimodal inputs is particularly challenging in this setting: data cannot leave each client, local samples are scarce and heterogeneously distributed, and unsupervised dimensionality reduction discards task-relevant information while federated training introduces representation drift across communication rounds. FDRMFL addresses these challenges through a unified four-term local objective: MSE prediction loss, a correlation-based mutual information surrogate that preserves dependence between the fused representation and the continuous target, a symmetric KL penalty that aligns cross-modal latent distributions before fusion, and an InfoNCE-style contrastive loss that anchors local representations to the global consensus. Experiments on three synthetic and two real-world near-infrared spectroscopy datasets under non-IID federated partitions, with comprehensive ablation and sensitivity analyses, demonstrate that each component contributes to the framework's effectiveness. FDRMFL reduces mean MSE by 33.8% relative to the best traditional baseline (PCA) and by 43.0% relative to VAE in simulation, and attains the lowest overall mean MSE among six federated algorithms including FedAvg, FedProx, MOON, SCAFFOLD, and FedBN.

Haozhe Wu

School of Computing and Data Science, The University of Hong Kong, Hong Kong SAR, China

Multimodal Federated Learning Non-IID Federated Learning Mutual Information Maximization Feature Fusion Federated Learning

Jun 15, 2025cs.CV

MGDFIS: Multi-scale Global-detail Feature Integration Strategy for Small Object Detection

Small-object detection in Unmanned Aerial Vehicle (UAV) imagery requires preserving weak local evidence while using broader context to separate tiny foreground targets from cluttered backgrounds. Existing multi-scale fusion methods improve feature aggregation, but they often add computation or blur fine details during repeated cross-scale fusion. The central challenge is to balance low-SNR target preservation, clutter suppression, and efficient cross-scale context exchange. To address this challenge, we propose the Multi-scale Global-detail Feature Integration Strategy (MGDFIS), a neck-level feature-fusion strategy that couples global context exchange, local-detail recovery, and pixel-level foreground-background recalibration. MGDFIS integrates three coordinated modules: FusionLock-TSS Attention for stabilizing spectral-spatial responses, Global-detail Integration for combining long-range mixing with local detail capture, and Dynamic Pixel Attention for reweighting compact foreground regions. On the controlled VisDrone setting, YOLO26m + MGDFIS improves AP50:95 from 25.7 to 30.2 and AP50 from 37.2 to 44.2 over the YOLO26m baseline, with 96.1 GFLOPs. Additional dataset-specific evaluations report 38.9 AP50 and 21.9 AP50:95 on UAVDT and 97.4 AP50 on CARPK. The code is available at: https://github.com/JackBaixue/MGDFIS.

Yuxiang Wang, Xuecheng Bai, Chuanzhi Xu +2

School of Computer Science, The University of Sydney, Sydney, Australia · 2Aviation Traffic Control Technology (SHENZHEN) Co., Ltd., Shenzhen, China

Remote Sensing Small Object Detection Small Object Detection Multi-Scale Feature Fusion Attention Mechanisms Aerial Object Detection Feature Fusion