Posted:October 16, 2019

Pulse: Deep Learning on Smartphones

AI3 Pulse

” . . . in the next two-three years all mid-range and high-end chipsets [on smartphones] will get enough power to run the vast majority of standard deep learning models developed by the research community and industry. This, in turn, will result in even more AI projects targeting mobile devices as the main platform for machine learning model deployment.”

The authors, most from leading smartphone providers, note AI is already used in selected smartphones:

“Among the most popular tasks are different computer vision problems like image classification, image enhancement, image super-resolution, bokeh simulation, object tracking, optical character recognition, face detection and recognition, augmented reality, etc. Another important group of tasks running on mobile devices is related to various NLP (Natural Language Processing) problems, such as natural language translation, sentence completion, sentence sentiment analysis, voice assistants and interactive chatbots. Additionally, many tasks deal with time series processing, e.g., human activity recognition, gesture recognition, sleep monitoring, adaptive power management, music tracking and classification.” (inline reference numbers removed)

Expect to see greater ubiquity and deeper applications.

Ignatov, A. et al. AI Benchmark: All About Deep Learning on Smartphones in 2019. arXiv:1910.06663 [cs] 1–19 (2019).

Schema.org Markup

headline:
Pulse: Deep Learning on Smartphones

alternativeHeadline:

author:

image:
https://www.mkbergman.com/wp-content/themes/ai3v2/images/pulse.png

description:
Smartphones are gaining rapidly in hardware capabilities such that most mid-range to upper-tier smartphones in the next few years will be able to run today's standard desktop AI tasks, resulting in powerful new mobile device applications.

articleBody:
see above

datePublished:

Leave a Reply

Your email address will not be published. Required fields are marked *