Gesture-SNN: Co-optimizing accuracy, latency and energy of SNNs for neuromorphic vision sensors

Singh, Sonali; Sarma, Anup; Lu, Sen; Sengupta, Abhronil; Narayanan, Vijaykrishnan; Das, Chita R.

doi:10.1109/islped52811.2021.9502506

Cited by 11 publications

(4 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We also highlight that the work outperforms computationally expensive BPTT based fine-tuning approaches since temporal information may not be relevant in static image classification tasks. Future exploration into application drivers with temporal information (Mahapatra et al, 2020 ; Singh et al, 2021 ) or temporal spike encoding schemes (Petro et al, 2020 ; Yang and Sengupta, 2020 ) is expected to truly leverage the full potential of BPTT based SNN training strategies.…”

Section: Discussionmentioning

confidence: 99%

Neuroevolution Guided Hybrid Spiking Neural Network Training

Sengupta

2022

Front. Neurosci.

Self Cite

View full text Add to dashboard Cite

Neuromorphic computing algorithms based on Spiking Neural Networks (SNNs) are evolving to be a disruptive technology driving machine learning research. The overarching goal of this work is to develop a structured algorithmic framework for SNN training that optimizes unique SNN-specific properties like neuron spiking threshold using neuroevolution as a feedback strategy. We provide extensive results for this hybrid bio-inspired training strategy and show that such a feedback-based learning approach leads to explainable neuromorphic systems that adapt to the specific underlying application. Our analysis reveals 53.8, 28.8, and 28.2% latency improvement for the neuroevolution-based SNN training strategy on CIFAR-10, CIFAR-100, and ImageNet datasets, respectively in contrast to state-of-the-art conversion based approaches. The proposed algorithm can be easily extended to other application domains like image classification in presence of adversarial attacks where 43.2 and 27.9% latency improvements were observed on CIFAR-10 and CIFAR-100 datasets, respectively.

show abstract

Section: Discussionmentioning

confidence: 99%

Neuroevolution Guided Hybrid Spiking Neural Network Training

Sengupta

2022

Front. Neurosci.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Additionally, as chips inspired by the form and functionality of biological neural circuits, such as Intel Loihi [ 258 ] are developed further, SNNs can enable learning in real‐time at low power levels. Additionally, SNNs are yielding good performance in various application areas where traditional Deep Learning approaches have faced considerable challenges, such as dynamic vision sensor applications [ 259 ] and smart grids [ 260 ] —where the sparse, temporal nature of the input data yields itself favorably to SNN algorithms.…”

Section: Anns and Hardware Aspectsmentioning

confidence: 99%

Complex Oxides for Brain‐Inspired Computing: A Review

et al. 2022

View full text Add to dashboard Cite

The fields of brain‐inspired computing, robotics, and, more broadly, artificial intelligence (AI) seek to implement knowledge gleaned from the natural world into human‐designed electronics and machines. In this review, the opportunities presented by complex oxides, a class of electronic ceramic materials whose properties can be elegantly tuned by doping, electron interactions, and a variety of external stimuli near room temperature, are discussed. The review begins with a discussion of natural intelligence at the elementary level in the nervous system, followed by collective intelligence and learning at the animal colony level mediated by social interactions. An important aspect highlighted is the vast spatial and temporal scales involved in learning and memory. The focus then turns to collective phenomena, such as metal‐to‐insulator transitions (MITs), ferroelectricity, and related examples, to highlight recent demonstrations of artificial neurons, synapses, and circuits and their learning. First‐principles theoretical treatments of the electronic structure, and in situ synchrotron spectroscopy of operating devices are then discussed. The implementation of the experimental characteristics into neural networks and algorithm design is then revewed. Finally, outstanding materials challenges that require a microscopic understanding of the physical mechanisms, which will be essential for advancing the frontiers of neuromorphic computing, are highlighted.

show abstract

“…1) Novel SNN Architectures: Earlier studies have demonstrated the effectiveness of SNNs in event-based object recognition and image classification tasks [45] [46], but their application in event-based object detection has been limited. However, in recent years, with the availability of large-scale event-based datasets, there has been a surge in research utilizing SNNs in the domain of event cameras.Building upon this foundation, Cordone [13] rebuilds effective object detection networks such as SqueezeNet [47], VGG [48], MobileNet [49], and DenseNet [50], leveraging recent advancements in spike-based backpropagation techniques like surrogate gradient learning, parameterized Leaky Integrate-and-Fire (LIF) neurons, and the SpikingJelly framework [51].…”

Section: Recurrent Vision Transformers For Object Detection With Even...mentioning

confidence: 99%

First cured patient with 2019 novel coronavirus in Changsha, China, a case report

Zhou

Jiang

Cao

et al. 2020

Preprint

View full text Add to dashboard Cite

Background: In December, 2019, the novel coronavirus (2019-nCoV) firstly occurred in Wuhan, Hubei Province, China. It spreads rapidly and many cases are identified in multiple countries, which pose a great threat to people's health. Here we report the first cured patient with 2019-nCoV infection in Changsha, China, and the symptoms, diagnosis, treatment and management of this patient are all described in the present study. Case presentation: A 57 year old woman developed cough and fever after returning from Wuhan to Changsha on January 9, 2020. She was tested positive for 2019-nCoV infection, which was supported by Chest CT. Lopinavir and ritonavir tablets and interferon alfa-2b injection were used for treatment. A small dose of glucocorticoids were used in a short period to control the immune response in bilateral lung and this patient avoided the occurrence of cytokine storms. The clinical condition of this patient improved and negative result was obtained for 2019-nCoV assay on January 25, 2020. This patient was recovered and discharged on January 30, 2020. Conclusions: At present, many reports about 2019-nCoV infection focused on the epidemiological and clinical characteristics of this disease. This case described the symptoms, diagnosis, treatment and management of a cured patient with 2019-nCoV infection, which may provide information for the treatment of 2019-nCoV infection.

show abstract

Gesture-SNN: Co-optimizing accuracy, latency and energy of SNNs for neuromorphic vision sensors

Cited by 11 publications

References 17 publications

Neuroevolution Guided Hybrid Spiking Neural Network Training

Neuroevolution Guided Hybrid Spiking Neural Network Training

Complex Oxides for Brain‐Inspired Computing: A Review

First cured patient with 2019 novel coronavirus in Changsha, China, a case report

Contact Info

Product

Resources

About