为了解决当前三阶段融合跟踪范式所面临的问题,本章提出了一个基于Transformer 的高效单阶段 RGB-T 单目标跟踪网络 USTrack(Unified Single-Stage Transformer Network for Efficient RGB-T Tracking)。其核心是通过联合特征提取 &融合 & 关联建模方法,将三阶段融合跟踪范式的三个功能部分直接统一到一个ViT主干网中同时...
为了解决当前三阶段融合跟踪范式所面临的问题,本章提出了一个基于Transformer 的高效单阶段 RGB-T 单目标跟踪网络 USTrack(Unified Single-Stage Transformer Network for Efficient RGB-T Tracking)。其核心是通过联合特征提取 &融合 & 关联建模方法,将三阶段融合跟踪范式的三个功能部分直接统一到一个ViT主干网中同时...
为了解决当前三阶段融合跟踪范式所面临的问题,本章提出了一个基于Transformer 的高效单阶段 RGB-T 单目标跟踪网络 USTrack(Unified Single-Stage Transformer Network for Efficient RGB-T Tracking)。其核心是通过联合特征提取 &融合 & 关联建模方法,将三阶段融合跟踪范式的三个功能部分直接统一到一个ViT主干网中同时...
RGB-Thermal(以下简称RGB-T)这个工作无论是在这篇文章中的Tracking,亦或是在人群计数,街景语义分割,夜景的目标检测中都有一个非常大的问题,就是缺数据。缺数据是目前RGB-T没有办法像RGB-D一样更为人知的原因之一,在缺少数据的情况下,整体训练的效果不尽如人意,我们也无法对于RGB-T进行进一步的探索和研究。这篇...
RGB-T trackingFeature fusionChannel communicationMany RGBT trackers utilize adaptive weighting mechanism to treat dual modalities differently and obtain more robust feature representations for tracking. Although these trackers work well under certain conditions, however, they ignore the information interactions...
常见的RGBT单目标跟踪数据集包括RGBT210、RGBT234、GTOT、VOT2019RGBT[80]、LasHeR(Large-scale High-diversity benchmark for RGBT tracking、VTUAV(Visible Thermal Unmanned Aerial Vehicle)等。 这些数据集为研究人员提供了一个标准化的平台,用于评估和比较不同单目标跟踪算法的性能。下表给出了上述数据集的详...
Cross Fusion RGB-T Tracking with Bi-directional Adapter 2024 6 AFter 87.663.5 AFter: Attention-based Fusion Router for RGBT Tracking 2024 7 MPLT 86.263.0 RGB-T Tracking via Multi-Modal Mutual Prompt Learning 2023 8 CSTNet 86.063.5
《Duality‑gated mutual condition network for rgbt tracking》使用注意力机制来学习每种模态对双重模态自适应融合的贡献,基于动态卷积 的RGBT跟踪器《Mfgnet:Dynamic modality‑aware filter generation for rgb‑t tracking》设计了动态的模态感知滤波器生成网络,通过自适应的调整卷积核来增强可见 光和热红外数据...
Fusion tracking based on visible and thermal infrared images can boost tracking performance under adverse challenging conditions, such as low illumination and bad weather. Existing RGBT tracking methods mainly focus on estimating the reliability weights of two modalities to achieve effective...
In recent years, many RGB-THERMAL tracking methods have been proposed to meet the needs of single object tracking under different conditions. However, these trackers are based on ANCHOR-BASED algorithms and feature cross-correlation operations, making it difficult to improve the success rate of targ...