.png){kind=icon}
.png){kind=icon}
Return
MPPM: A Mobile-Efficient Part Model for Object re-ID
.png){kind=icon}
PRE
AI
DOI：10.1109/TMM.2022.3207895.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Object re-identification (re-ID) is one of the core technologies in Multi-Object Tracking (MOT) that requires realtime decision-making. A Neural Processing Unit (NPU) is a lowpower device that is dedicated to deploying neural network-based algorithms and has become one of the most important devices in today'smobile onboard systems. However, the currentmainstream re-ID methods rarely consider the NPU characteristics, which makes it difficult for these methods to achieve both high onboard frame rates and high accuracies on an NPU. To address this problem, this paper focuses on designing a re-ID algorithm suitable for NPU deployment. The model of the object re-ID can be divided into two parts: the encoder (backbone) and the decoder. In this article, a Mobile-efficient Pure Part Model (MPPM) is presented for re-ID task. First, for the backbone of re-ID, we propose an efficient structure GogglesNet, which is composed of traditional convolutions. GogglesNet performs well on the re-ID task and can be comparable to lightweight networks on ImageNet with regard to accuracy and is faster on NPU. We then revisit the architectures of Pure Part Model (PPM) in person re-ID, including PCB and MGN, and propose a mobile-efficient decoder Dual Pattern Network (DPN) for re-ID. The proposed MPPM achieves comparable performance with MGN on five reID datasets Market-1501, DukeMTMC-reID, MSMT17, VeRi-776, andVehicleID, while the proposed parameter amount is only 10.2% of it, and the speed on NPU is more than eight times higher.
Keywords:
Task analysis
Decoding
Hardware
Neural networks
Eye protection
Electronic mail
Costs
Mobile-efficient
neural processing units
object re-ID
part model
Journal
IF:
9.7
Papers:
4.5K
Citations:
2.4W