Target-Specific Siamese Attention Network for Real-Time Object Tracking

Thanikasalam, Kokul; Fookes, Clinton; Sridharan, Sridha; Ramanan, Amirthalingam; Pinidiyaarachchi, Amalka

doi:10.1109/tifs.2019.2935871

Cited by 12 publications

(3 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Huang et al [15] proposed an attentional online update paradigm for siamese visual tracking to improve the performance of a tracker by utilizing knowledge extracted from prior tracking tasks. In [32] residual attention modules are introduced in similarity tracking at multiple levels of feature representation, resulting in improved discrimination quality for similarity searching. Zhang et al [40] created an attention retrieval network that uses learning masks to conduct soft spatial constrains on features from a tracking backbone network, mitigating the impact of background clutter.…”

Section: Attention Mechanisms In Object Trackingmentioning

confidence: 99%

AttTrack: Online Deep Attention Transfer for Multi-object Tracking

Nalaie¹,

Zheng²

2022

Preprint

View full text Add to dashboard Cite

Multi-object tracking (MOT) is a vital component of intelligent video analytics applications such as surveillance and autonomous driving. The time and storage complexity required to execute deep learning models for visual object tracking hinder their adoption on embedded devices with limited compute power. In this paper, we aim to accelerate MOT by transferring the knowledge from high-level features of a complex network (teacher) to a lightweight network (student) at both training and inference time. The proposed AttTrack framework has three key components: 1) crossmodel feature learning to align intermediate representations from the teacher and student models, 2) interleaving the execution of the two models at inference time, and 3) incorporating the updated predictions from the teacher model as prior knowledge to assist the student model. Experiments on pedestrian tracking tasks are conducted on the MOT17 and MOT15 datasets using two different object detection backbones YOLOv5 and DLA34 show that AttTrack can significantly improve student model tracking performance while sacrificing only minor degradation of tracking speed.

show abstract

Section: Attention Mechanisms In Object Trackingmentioning

confidence: 99%

AttTrack: Online Deep Attention Transfer for Multi-object Tracking

Nalaie¹,

Zheng²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Attention mechanism is similar to human's visual attention: we always focus on the most important part of what we see. It has been widely utilized in many other fields, such as NLP [32,33] and CV [34,35]. Specifically, attention mechanism makes it easy to memorize various remote dependencies or focus on important parts of the input.…”

Section: Attention Mechanismmentioning

confidence: 99%

TLSAN: Time-aware Long- and Short-term Attention Network for Next-item Recommendation

Zhang,

Wang,

2021

Preprint

View full text Add to dashboard Cite

Recently, deep neural networks are widely applied in recommender systems for their effectiveness in capturing/modeling users' preferences. Especially, the attention mechanism in deep learning enables recommender systems to incorporate various features in an adaptive way. Specifically, as for the next item recommendation task, we have the following three observations: 1) users' sequential behavior records aggregate at time positions ("time-aggregation"), 2) users have personalized taste that is related to the "time-aggregation" phenomenon ("personalized time-aggregation"), and 3) users' short-term interests play an important role in the next item prediction/recommendation. In this paper, we propose a new Time-aware Long-and Short-term Attention Network (TLSAN) to address those observations mentioned above. Specifically, TLSAN consists of two main components. Firstly, TLSAN models "personalized time-aggregation" and learn user-specific temporal taste via trainable personalized time position embeddings with category-aware correlations in long-term behaviors. Secondly, long-and short-term feature-wise attention layers are proposed to effectively capture users' long-and short-term preferences for accurate recommendation. Especially, the attention mechanism enables TLSAN to utilize users' preferences in an adaptive way, and its usage in long-and short-term layers enhances TLSAN's ability of dealing with sparse interaction data. Extensive experiments are conducted on Amazon datasets from different fields (also with different size), and the results show that TLSAN outperforms state-ofthe-art baselines in both capturing users' preferences and performing time- * Corresponding author.

show abstract

“…Compared to the recent Siamese based tracker [44], the performance of our tracker is superior to [44] (We use the reported results for fair comparisons.) in terms of distance precision (DP: 91.1% vs. 87.7%) and overlap success rate (OS AUC : 66.6% vs. 66.4%).…”

Section: ) Otb-2015 Datasetmentioning

confidence: 99%

Learning Objectness Transfer Networks for Visual Tracking

Zhao

Zhang

2019

IEEE Access

View full text Add to dashboard Cite

Existing deep trackers mainly use deep neural networks pre-trained on the object recognition training sets to generate deep features as target representation. However, pre-trained deep features are not effective in representing arbitrary forms of target objects which are likely to be unseen for the pretrained deep networks. To narrow the gap of representation capability, we propose to transfer the objectness information within pre-trained deep networks. The transferred objectness information is utilized to generate deep features aware of any arbitrary form of target objects for robust visual tracking. Specifically, we design a novel network branch on top of pre-trained deep models to perform incremental transfer learning. The learned network with the transferred objectness information helps to locate target objects undergoing large appearance changes precisely. Experimental results on standard benchmark datasets demonstrate that the proposed algorithm performs favorably against the start-of-the-art trackers.INDEX TERMS Visual tracking, transfer learning, objectness, deep features.

show abstract

Target-Specific Siamese Attention Network for Real-Time Object Tracking

Cited by 12 publications

References 57 publications

AttTrack: Online Deep Attention Transfer for Multi-object Tracking

AttTrack: Online Deep Attention Transfer for Multi-object Tracking

TLSAN: Time-aware Long- and Short-term Attention Network for Next-item Recommendation

Learning Objectness Transfer Networks for Visual Tracking

Contact Info

Product

Resources

About