High-Capacity Expert Binary Networks

• URL: http://arxiv.org/abs/2010.03558v2
• Date: Tue, 30 Mar 2021 18:16:16 GMT
• Title: High-Capacity Expert Binary Networks
• Authors: Adrian Bulat and Brais Martinez and Georgios Tzimiropoulos
• Abstract summary: Network binarization is a promising hardware-aware direction for creating efficient deep models. Despite its memory and computational advantages, reducing the accuracy gap between binary models and their real-valued counterparts remains an unsolved challenging research problem. We propose Expert Binary Convolution, which, for the first time, tailors conditional computing to binary networks by learning to select one data-specific expert binary filter at a time conditioned on input features.
• Score: 56.87581500474093
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Network binarization is a promising hardware-aware direction for creating efficient deep models. Despite its memory and computational advantages, reducing the accuracy gap between binary models and their real-valued counterparts remains an unsolved challenging research problem. To this end, we make the following 3 contributions: (a) To increase model capacity, we propose Expert Binary Convolution, which, for the first time, tailors conditional computing to binary networks by learning to select one data-specific expert binary filter at a time conditioned on input features. (b) To increase representation capacity, we propose to address the inherent information bottleneck in binary networks by introducing an efficient width expansion mechanism which keeps the binary operations within the same budget. (c) To improve network design, we propose a principled binary network growth mechanism that unveils a set of network topologies of favorable properties. Overall, our method improves upon prior work, with no increase in computational cost, by $\sim6 \%$, reaching a groundbreaking $\sim 71\%$ on ImageNet classification. Code will be made available $\href{https://www.adrianbulat.com/binary-networks}{here}$.

Related papers

• BiHRNet: A Binary high-resolution network for Human Pose Estimation [11.250422970707415]
  We propose a binary human pose estimator named BiHRNet, whose weights and activations are expressed as $pm$1. BiHRNet retains the keypoint extraction ability of HRNet, while using fewer computing resources by adapting binary neural network (BNN) We show BiHRNet achieves a PCKh of 87.9 on the MPII dataset, which outperforms all binary pose estimation networks.
  arXiv  Detail & Related papers  (2023-11-17T03:01:37Z)
• Input Layer Binarization with Bit-Plane Encoding [4.872439392746007]
  We present a new method to binarize the first layer using directly the 8-bit representation of input data. The resulting model is fully binarized and our first layer binarization approach is model independent.
  arXiv  Detail & Related papers  (2023-05-04T14:49:07Z)
• BiBench: Benchmarking and Analyzing Network Binarization [72.59760752906757]
  Network binarization emerges as one of the most promising compression approaches offering extraordinary computation and memory savings. Common challenges of binarization, such as accuracy degradation and efficiency limitation, suggest that its attributes are not fully understood. We present BiBench, a rigorously designed benchmark with in-depth analysis for network binarization.
  arXiv  Detail & Related papers  (2023-01-26T17:17:16Z)
• Join the High Accuracy Club on ImageNet with A Binary Neural Network Ticket [10.552465253379134]
  We focus on a problem: how can a binary neural network achieve the crucial accuracy level (e.g., 80%) on ILSVRC-2012 ImageNet? We design a novel binary architecture BNext based on a comprehensive study of binary architectures and their optimization process. We propose a novel knowledge-distillation technique to alleviate the counter-intuitive overfitting problem observed when attempting to train extremely accurate binary models.
  arXiv  Detail & Related papers  (2022-11-23T13:08:58Z)
• BiFSMNv2: Pushing Binary Neural Networks for Keyword Spotting to Real-Network Performance [54.214426436283134]
  Deep neural networks, such as the Deep-FSMN, have been widely studied for keyword spotting (KWS) applications. We present a strong yet efficient binary neural network for KWS, namely BiFSMNv2, pushing it to the real-network accuracy performance. We highlight that benefiting from the compact architecture and optimized hardware kernel, BiFSMNv2 can achieve an impressive 25.1x speedup and 20.2x storage-saving on edge hardware.
  arXiv  Detail & Related papers  (2022-11-13T18:31:45Z)
• Robust Training and Verification of Implicit Neural Networks: A Non-Euclidean Contractive Approach [64.23331120621118]
  This paper proposes a theoretical and computational framework for training and robustness verification of implicit neural networks. We introduce a related embedded network and show that the embedded network can be used to provide an $ell_infty$-norm box over-approximation of the reachable sets of the original network. We apply our algorithms to train implicit neural networks on the MNIST dataset and compare the robustness of our models with the models trained via existing approaches in the literature.
  arXiv  Detail & Related papers  (2022-08-08T03:13:24Z)
• Binarizing MobileNet via Evolution-based Searching [66.94247681870125]
  We propose a use of evolutionary search to facilitate the construction and training scheme when binarizing MobileNet. Inspired by one-shot architecture search frameworks, we manipulate the idea of group convolution to design efficient 1-Bit Convolutional Neural Networks (CNNs) Our objective is to come up with a tiny yet efficient binary neural architecture by exploring the best candidates of the group convolution.
  arXiv  Detail & Related papers  (2020-05-13T13:25:51Z)
• ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions [76.05981545084738]
  We propose several ideas for enhancing a binary network to close its accuracy gap from real-valued networks without incurring any additional computational cost. We first construct a baseline network by modifying and binarizing a compact real-valued network with parameter-free shortcuts. We show that the proposed ReActNet outperforms all the state-of-the-arts by a large margin.
  arXiv  Detail & Related papers  (2020-03-07T02:12:02Z)
• Exploring the Connection Between Binary and Spiking Neural Networks [1.329054857829016]
  We bridge the recent algorithmic progress in training Binary Neural Networks and Spiking Neural Networks. We show that training Spiking Neural Networks in the extreme quantization regime results in near full precision accuracies on large-scale datasets.
  arXiv  Detail & Related papers  (2020-02-24T03:46:51Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.