Training Competitive Binary Neural Networks from Scratch

Abstract

Convolutional neural networks have achieved astonishing results in different application areas. Various methods that allow us to use these models on mobile and embedded devices have been proposed. Especially binary neural networks are a promising approach for devices with low computational power. However, training accurate binary models from scratch remains a challenge. Previous work often uses prior knowledge from full-precision models and complex training strategies. In our work, we focus on increasing the performance of binary neural networks without such prior knowledge and a much simpler training strategy. In our experiments we show that we are able to achieve state-of-the-art results on standard benchmark datasets. Further, to the best of our knowledge, we are the first to successfully adopt a network architecture with dense connections for binary networks, which lets us improve the state-of-the-art even further.

Our source code can be found at https://github.com/hpi-xnor/BMXNet-v2

References in corpus (9)

• Adam: A Method for Stochastic Optimization
• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
• Deep Residual Learning for Image Recognition
• Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding
• DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients
• Binarized Neural Networks
• Towards Accurate Binary Convolutional Neural Network
• Bi-Real Net: Enhancing the Performance of 1-bit CNNs With Improved Representational Capability and Advanced Training Algorithm

Cited by in corpus (15)

• Binary Neural Networks: A Survey
• XNOR-Net++: Improved Binary Neural Networks
• MeliusNet: Can Binary Neural Networks Achieve MobileNet-level Accuracy?
• Benchmarking Quantized Neural Networks on FPGAs with FINN
• FxP-QNet: A Post-Training Quantizer for the Design of Mixed Low-Precision DNNs with Dynamic Fixed-Point Representation
• Effective Training of Convolutional Neural Networks with Low-bitwidth Weights and Activations
• Towards Learning of Filter-Level Heterogeneous Compression of Convolutional Neural Networks
• Mixed-Precision Quantized Neural Network with Progressively Decreasing Bitwidth For Image Classification and Object Detection
• BARS: Joint Search of Cell Topology and Layout for Accurate and Efficient Binary ARchitectures
• Binarizing MobileNet via Evolution-based Searching
• Path Sample-Analytic Gradient Estimators for Stochastic Binary Networks
• Training Quantized Neural Networks with a Full-precision Auxiliary Module
• BCNN: Binary Complex Neural Network
• FATNN: Fast and Accurate Ternary Neural Networks
• Robustness-aware 2-bit quantization with real-time performance for neural network