期刊名称: Information Fusion 论文地址: 10.1109/MVIPIT60427.2023.00035 作者: 单位: 北京大学计算机学院 1.研究目的 该论文的主要研究目的是开发一种基于变压器的多模态跨尺度特征融合网络,用于提高复杂环境下车辆检测的准确性和鲁棒性。为了应对在不同环境(如白天与夜晚、晴天与雾天)下的车辆检测挑战,作者引入了多...
In this paper, we propose an cross-scale feature fusion network (CFFNet) to harvest the compact segmentatiHon model with high accuracy. Specifically, we design a novel lightweight residual block in backbone with increasing block depth strategy instead of inverted residual block with increasing ...
feature fusion of different scales, effectively bridging the semantic gap between high-resolution features and low-resolution features. An adaptive feature fusion module (AFFM) is proposed, which adaptively fuses high-resolution and low-resolution features, which helps to improve the accuracy of ...
Secondly, the Neck introduces a novel architecture to enhance feature fusion by integrating the RepBlock and BiFusion modules. Finally, the performance of the fish detection model is demonstrated based on the Fish26 dataset, in which the detection accuracy, computational cost, and parameter ...
First, a cross-scale feature fusion module is devised to improve the detection performance of multiscale and partially occluded objects. This module fuses the global features of different residual layers and multiscale local region features, thereby improving the multiscale feature fusion capability of...
More specifically, to effectively combine the information from multiple references, a cross-scale and cross-space feature fusion mechanism is introduced for external and internal references, and then the enhanced local semantics are finally incorporated into the high-resolution face reconstruction. The ...
It should be noted that the feature sizes of transverse connections are the same. As shown in Figure 5, the red arrows represent the output branches, the blue arrows represent the correction branches, and the yellow circles indicate the Cross-Level Feature Fusion Module (CLFFM). The correction...
MFCPNet: Real time medical image segmentation network via multi-scale feature fusion and channel pruning Real-time medical image segmentation can not only enhance the interactivity and feasibility of applications but also support more medical application scena... Linlin Hou,Zishen Yan,Christian ...
Object detection usually adopts multi-scale fusion to enrich the information of the object, and the Feature Pyramid Network (FPN) is a common method for multi-scale fusion. However, traditional fusion methods such as FPN cause information loss when fusing high-level feature maps with low-level ...
Finally, we utilize the Feature Pyramid Aggregation Network for multi-scale feature fusion, achieving a deeper integration of the complementary data. Extensive experiments show that the proposed approach achieves competitive performances on MP6D, LM, and LM-O datasets....