System Demo for Transfer Learning from Vision to Language using Domain Specific CNN Accelerator for On-Device NLP Applications

Anonymous

Keywords: Interactive System Demonstration, Text Classification, Sentiment Analysis, Transfer Learning, AI Chip, Two-dimensional Word Embedding

Abstract: Power-efficient CNN Domain Specific Accelerator (CNN-DSA) chips are currently available for wide use in mobile devices. These chips are mainly used in computer vision applications. However, the recent work of Super Characters method for text classification and sentiment analysis tasks using two-dimensional CNN models has also achieved state-of-the-art results through the method of transfer learning from vision to text. In this paper, we implemented the text classification and sentiment analysis applications on mobile devices using CNN-DSA chips. Compact network representations using one-bit and three-bits precision for coefficients and five-bits for activations are used in the CNN-DSA chip with power consumption less than 300mW. For edge devices under memory and compute constraints, the network is further compressed by approximating the external Fully Connected (FC) layers within the CNN-DSA chip. At the workshop, we have two system demonstrations for NLP tasks. The first demo classifies the input English Wikipedia sentence into one of the 14 classes. The second demo classifies the Chinese online-shopping review into positive or negative.

TL;DR: Deploy text classification and sentiment analysis applications for English and Chinese on a 300mW CNN accelerator chip for on-device application scenarios.