检索结果-内蒙古大学图书馆

32nd ACM International Conference on Information and Knowledge Management (CIKM)

作者： Liu, Qi Zhou, Zhilong Jiang, Gangwei Ge, Tiezheng Lian, Defu Univ Sci & Technol China Hefei Peoples R China Alibaba Grp Hangzhou Peoples R China

ISBN: (纸本)9798400701245

Neural-based multi-task learning (MTL) has gained significant improvement, and it has been successfully applied to recommendation system (RS). Recent deep MTL methods for RS (e.g. MMoE, PLE) focus on designing soft gating-based parameter-sharing networks that implicitly learn a generalized representation for each task. However, MTL methods may suffer from performance degeneration when dealing with conflicting tasks, as negative transfer effects can occur on the task-shared bottom representation. This can result in a reduced capacity for MTL methods to capture task-specific characteristics, ultimately impeding their effectiveness and hindering the ability to generalize well on all tasks. In this paper, we focus on the bottom representation learning of MTL in RS and propose the Deep task-specific Bottom representation Network (DTRN) to alleviate the negative transfer problem. DTRN obtains task-specific bottom representation explicitly by making each task have its own representation learning network in the bottom representation modeling stage. specifically, it extracts the user's interests from multiple types of behavior sequences for each task through the parameter-efficient hypernetwork. To further obtain the dedicated representation for each task, DTRN refines the representation of each feature by employing a SENet-like network for each task. The two proposed modules can achieve the purpose of getting task-specific bottom representation to relieve tasks' mutual interference. Moreover, the proposed DTRN is flexible to combine with existing MTL methods. Experiments on one public dataset and one industrial dataset demonstrate the effectiveness of the proposed DTRN.

关键词： multi-task recommendation hypernetwork task-specific representation behavior sequence modeling

来源：评论

学校读者我要写书评

暂无评论

Learning task-specific Morphological representation for Pyramidal Cells via Mutual Information Minimization 6th

Learning Task-Specific Morphological Representation for Pyra...

引用

6th International Workshop on Predictive Intelligence in Medicine (PRIME)

作者： Sun, Chunli Guo, Qinghai Yang, Gang Zhao, Feng Univ Sci & Technol China Hefei Peoples R China ACS Lab Huawei Technol Shenzhen Peoples R China

ISBN: (纸本)9783031460043;9783031460050

The morphology of pyramidal cells (PCs) varies significantly among species and brain layers. Therefore, it is particularly challenging to analyze which species or layers they belong to based on morphological features. Existing deep learning-based methods analyze speciesrelated or layer-related morphological characteristics of PCs. However, these methods are realized in a task-agnostic manner without considering task-specific features. This paper proposes a task-specific morphological representation learning framework for morphology analysis of PCs to enforce task-specific feature extraction through dual-task learning, enabling performance gains for each task. specifically, we first utilize species-wise and layer-wise feature extraction branches to obtain species-related and layer-related features. Applying the principle of mutual information minimization, we then explicitly force each branch to learn taskspecific features, which are further enhanced via an adaptive representation enhancement module. In this way, the performance of both tasks can be greatly improved simultaneously. Experimental results demonstrate that the proposed method can effectively extract the species-specific and layer-specific representations when identifying rat and mouse PCs in multiple brain layers. Our method reaches the accuracies of 87.44% and 72.46% on species and layer analysis tasks, significantly outperforming a single task by 2.22% and 3.86%, respectively.

关键词： Mutual information Morphology analysis Type prediction task-specific representation

来源：评论

学校读者我要写书评

暂无评论

Learning task-specific discriminative representations for multiple object tracking

引用

NEURAL COMPUTING & APPLICATIONS 2023年第10期35卷 7761-7777页

作者： Wu, Han Nie, Jiahao Zhu, Ziming He, Zhiwei Gao, Mingyu Hangzhou Dianzi Univ Sch Elect Informat Hangzhou 310018 Zhejiang Peoples R China Zhejiang Prov Key Lab Equipment Elect 2019E10009 Hangzhou Zhejiang Peoples R China

One-shot multiple object tracking (MOT), which learns object detection and identity embedding in a unified network, has attracted increasing attention due to its low complexity and high tracking speed. However, most one-shot trackers ignore that detection and re-identification (ReID) require different representations of features. The inherent difference between these two subtasks leads to optimization contradictions in the training procedure. This issue would result in suboptimal tracking performance. To alleviate this contradiction, we propose a novel dual-path transformation network (DTN) that decouples the shared features into detection-specific and ReID-specific representations. By learning task-specific features, this module satisfies the different requirements of both subtasks. Moreover, we observe that previous trackers generally utilize local information to distinguish targets and ignore global semantic relations, which are crucial for tracking. Therefore, we design a pyramid non-local network (PNN) that allows our network to explore pixel-to-pixel relations with a global receptive field. Meanwhile, PNN considers the scale information to enhance the robustness to scale variations. Extensive experiments conducted on three benchmarks, i.e., MOT16, MOT17, and MOT20, demonstrate the superiority of our tracker, namely DPTrack. The experimental results reveal that DPTrack achieves state-of-the-art performance, e.g., MOTA of 77.1% on MOT17. Moreover, DPTrack runs at 14.9FPS, and our lightweight version runs at 26.6FPS with only a slight performance decay.

关键词： Multiple object tracking task-specific representation Optimization contradiction Scare-aware discriminative embedding

来源：评论

学校读者我要写书评

暂无评论

One-Shot Multiple Object Tracking in UAV Videos Using task-specific Fine-Grained Features

引用

REMOTE SENSING 2022年第16期14卷 3853-3853页

作者： Wu, Han Nie, Jiahao He, Zhiwei Zhu, Ziming Gao, Mingyu Hangzhou Dianzi Univ Sch Elect Informat Hangzhou 310018 Peoples R China Zhejiang Prov Key Lab Equipment Elect Hangzhou 310009 Peoples R China

Multiple object tracking (MOT) in unmanned aerial vehicle (UAV) videos is a fundamental task and can be applied in many fields. MOT consists of two critical procedures, i.e., object detection and re-identification (ReID). One-shot MOT, which incorporates detection and ReID in a unified network, has gained attention due to its fast inference speed. It significantly reduces the computational overhead by making two subtasks share features. However, most existing one-shot trackers struggle to achieve robust tracking in UAV videos. We observe that the essential difference between detection and ReID leads to an optimization contradiction within one-shot networks. To alleviate this contradiction, we propose a novel feature decoupling network (FDN) to convert shared features into detection-specific and ReID-specific representations. The FDN searches for characteristics and commonalities between the two tasks to synergize detection and ReID. In addition, existing one-shot trackers struggle to locate small targets in UAV videos. Therefore, we design a pyramid transformer encoder (PTE) to enrich the semantic information of the resulting detection-specific representations. By learning scale-aware fine-grained features, the PTE empowers our tracker to locate targets in UAV videos accurately. Extensive experiments on VisDrone2021 and UAVDT benchmarks demonstrate that our tracker achieves state-of-the-art tracking performance.

关键词： multiple object tracking unmanned aerial vehicle video optimization contradiction task-specific representation transformer encoder fine-grained feature

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：