A Low-Power Deep Neural Network Online Learning Processor for Real-Time Object Tracking Application

Han, Donghyeon; Lee, Jinsu; Lee, Jinmook; Yoo, Hoi‐Jun

doi:10.1109/tcsi.2018.2880363

Cited by 45 publications

(36 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A HW accelerator close to our work was recently published in [14]. The authors implement a HW accelerator in 65nm ASIC to train the last fully connected layers in the MDNET based object tracker similar to our work.…”

Section: Relu and Dropout Storagementioning

confidence: 95%

“…Our HW accelerator can be adopted to train other Multi-Layer Perception (MLP) neural networks that are employed in real-time applications on embedded devices. Some prior works [14], [23], [24] exist in literature to accelerate the training of the MLP or fully-connected neural networks. Approximate computing is adopted in [23] via inexact multipliers and bit-precision reduction to reduce the power consumption.…”

Section: Relu and Dropout Storagementioning

confidence: 99%

“…Moreover, a fixed-point hardware accelerator is developed for the online-training computations. In general, CNN training requires at least 5x more multiply and accumulate (MAC) operations than the inference phase [14] with much larger amount of external memory accesses. These requirements challenge the real-time operation on embedded systems.…”

mentioning

confidence: 99%

See 2 more Smart Citations

An Efficient Hardware Implementation of CNN-Based Object Trackers for Real-Time Applications

El-Shafie¹,

Zaki²,

Habib³

2021

Preprint

View full text Add to dashboard Cite

<div>The object tracking research continues to be active since long period because of the several real-world variations imposed in the tracking process, like occlusion, changing appearance, illumination changes and cluttered background. With wide range of applications, embedded implementations are typically pursed for the tracking systems. Although object trackers based on Convolution Neural Network (CNN) have achieved state-of-the-art performance, they challenge the embedded implementations because of slow speed and large memory requirement. In this paper, we address these limitations on the algorithm-side and the circuitside. On the algorithm side, we adopt interpolation schemes which can significantly reduce the processing time and the memory storage requirements. We also evaluate the approximation of the hardware-expensive computations aiming for an efficient hardware implementation. Moreover, we modified the online-training scheme in order to achieve a constant processing time across all video frames. On the circuit side, we developed a hardware accelerator of the online training stage. We avoid the transposed reading from the external memory to speed-up the data movement with no performance degradation. Our proposed hardware accelerator achieves 45.9 frames-per-second in training the fully connected layers.</div>

show abstract

Section: Relu and Dropout Storagementioning

confidence: 95%

Section: Relu and Dropout Storagementioning

confidence: 99%

See 1 more Smart Citation

An Efficient Hardware Implementation of CNN-Based Object Trackers for Real-Time Applications

El-Shafie¹,

Zaki²,

Habib³

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Han et al design a dedicated processor for DNN-based real-time object tracking [70]. is processor achieves low power consumption through a DNN-specific processor architecture and a specialized algorithm.…”

Section: Hardware Revolutionmentioning

confidence: 99%

Deep Learning on Computational-Resource-Limited Platforms: A Survey

Chen

Zhang

et al. 2020

Mobile Information Systems

View full text Add to dashboard Cite

Nowadays, Internet of Things (IoT) gives rise to a huge amount of data. IoT nodes equipped with smart sensors can immediately extract meaningful knowledge from the data through machine learning technologies. Deep learning (DL) is constantly contributing significant progress in smart sensing due to its dramatic superiorities over traditional machine learning. The promising prospect of wide-range applications puts forwards demands on the ubiquitous deployment of DL under various contexts. As a result, performing DL on mobile or embedded platforms is becoming a common requirement. Nevertheless, a typical DL application can easily exhaust an embedded or mobile device owing to a large amount of multiply and accumulate (MAC) operations and memory access operations. Consequently, it is a challenging task to bridge the gap between deep learning and resource-limited platforms. We summarize typical applications of resource-limited deep learning and point out that deep learning is an indispensable impetus of pervasive computing. Subsequently, we explore the underlying reasons for the high computational overhead of DL through reviewing the fundamental concepts including capacity, generalization, and backpropagation of a neural network. Guided by these concepts, we investigate on principles of representative research works, as well as three types of solutions: algorithmic design, computational optimization, and hardware revolution. In pursuant to these solutions, we identify challenges to be addressed.

show abstract

“…The large number of multi-bit elementary arithmetic operations performed by the vector-matrix multiplier (VMM) and the heavy data exchange between the memory and logic elements represent the main limiting factors of both EE and throughput in conventional digital CPU architectures [1], [2], [3], [4]. The recurring nature of these arithmetic operations can be exploited by taking advantage of the parallel computing capability of GPUs [5] and of embedded ASIC accelerators [6], [7]. Parallelism in computation and in memory access can be better exploited through in-memory computing architectures, consisting of a large number of modularized processing elements distributed in space and operating in parallel, where each processing element contains both the logic and the memory to perform the assigned partial processing task.…”

Section: Introductionmentioning

confidence: 99%

Analog Vector-Matrix Multiplier Based on Programmable Current Mirrors for Neural Network Integrated Circuits

et al. 2020

View full text Add to dashboard Cite

A Low-Power Deep Neural Network Online Learning Processor for Real-Time Object Tracking Application

Cited by 45 publications

References 23 publications

An Efficient Hardware Implementation of CNN-Based Object Trackers for Real-Time Applications

An Efficient Hardware Implementation of CNN-Based Object Trackers for Real-Time Applications

Deep Learning on Computational-Resource-Limited Platforms: A Survey

Analog Vector-Matrix Multiplier Based on Programmable Current Mirrors for Neural Network Integrated Circuits

Contact Info

Product

Resources

About