DFTerNet: Towards 2-bit Dynamic Fusion Networks for Accurate Human Activity Recognition
Authors:
- Zhan YangZYZhan Yang
- Osolo Ian RaymondOROsolo Ian Raymond
- ChengYuan ZhangCZChengYuan Zhang
- Ying WanYWYing Wan
- Jun LongJLJun Long
Preprint
Abstract
Deep Convolutional Neural Networks (DCNNs) are currently popular in human activity recognition applications. However, in the face of modern artificial intelligence sensor-based games, many research achievements cannot be practically applied on portable devices. DCNNs are typically resource-intensive and too large to be deployed on portable devices, thus this limits the practical application of complex activity detection. In addition, since portable devices do not possess high-performance Graphic Processing Units (GPUs), there is hardly any improvement in Action Game (ACT) experience. Besides, in order to deal with multi-sensor collaboration, all previous human activity recognition models typically treated the representations from different sensor signal sources equally. However, distinct types of activities should adopt different fusion strategies. In this paper, a novel scheme is proposed. This scheme is used to train 2-bit Convolutional Neural Networks with weights and activations constrained to {-0.5,0,0.5}. It takes into account the correlation between different sensor signal sources and the activity types. This model, which we refer to as DFTerNet, aims at producing a more reliable inference and better trade-offs for practical applications. Our basic idea is to exploit quantization of weights and activations directly in pre-trained filter banks and adopt dynamic fusion strategies for different activity types. Experiments demonstrate that by using dynamic fusion strategy can exceed the baseline model performance by up to ~5% on activity recognition like OPPORTUNITY and PAMAP2 datasets. Using the quantization method proposed, we were able to achieve performances closer to that of full-precision counterpart. These results were also verified using the UniMiB-SHAR dataset. In addition, the proposed method can achieve ~9x acceleration on CPUs and ~11x memory saving.Keywords
All Related Versions
This publication has 16 references indexed in Scilit:
- CS-CNN: Enabling Robust and Efficient Convolutional Neural Networks Inference for Internet-of-Things ApplicationsIEEE Access, 2018
- Facial Expression Recognition Using Salient Features and Convolutional Neural NetworkIEEE Access, 2017
- G-CNN: Object Detection via Grid Convolutional Neural NetworkIEEE Access, 2017
- Gaze-Assisted Multi-Stream Deep Neural Network for Action RecognitionIEEE Access, 2017
- Faster Population Counts Using AVX2 InstructionsThe Computer Journal, 2017
- ECG Monitoring System Integrated with IR-UWB Radar Based on CNNIEEE Access, 2016
- Convolutional networks for fast, energy-efficient neuromorphic computingProceedings of the National Academy of Sciences, 2016
- XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks2016
- The Opportunity challenge: A benchmark database for on-body sensor-based activity recognitionPattern Recognition Letters, 2013
- Training binary node feedforward neural networks by back propagation of errorElectronics Letters, 1990