Accelerating Very Deep Convolutional Networks for Classification and Detection

Abstract

Abstract-This paper aims to accelerate the test-time computation of convolutional neural networks (CNNs), especially very deep CNNs [1] that have substantially impacted the computer vision community. Unlike previous methods that are designed for approximating linear filters or linear responses, our method takes the nonlinear units into account. We develop an effective solution to the resulting nonlinear optimization problem without the need of stochastic gradient descent (SGD). More importantly, while previous methods mainly focus on optimizing one or two layers, our nonlinear method enables an asymmetric reconstruction that reduces the rapidly accumulated error when multiple (e.g., ≥10) layers are approximated. For the widely used very deep VGG-16 model [1], our method achieves a whole-model speedup of 4× with merely a 0.3% increase of top-5 error in ImageNet classification. Our 4× accelerated VGG-16 model also shows a graceful accuracy degradation for object detection when plugged into the Fast R-CNN detector [2].

Quantized Convolutional Neural Networks for Mobile Devices

“…• CPD [17]: CP-Decomposition; • GBD [18]: Group-wise Brain Damage; • LANR [31]: Low-rank Approximation of Non-linear Responses.…”

“…In Table 5, we also report the comparison against LANR [31] on VGG-16. For the similar speed-up rate (4×), their approach outperforms ours in the top-5 classification error (an increase of 0.95% against 1.83%).…”

Learning a Dilated Residual Network for SAR Image Despeckling

“…Compared with traditional data mapping, He et al [35] found that residual mapping can acquire a more effective learning effect and rapidly reduce the training loss after passing through a multi-layer network, which has achieved a state-of-the-art performance in object detection [36], image super-resolution [37], and so on. Essentially, Szegedy et al [38] demonstrated that residual networks take full advantage of identity shortcut connections, which can efficiently transfer various levels of feature information between not directly connected layers without attenuation.…”

Reducing Smartwatch Users’ Distraction with Convolutional Neural Network

“…Examples include the CNN, widely used for image classi cation [25,26] and recently for text classi cation [27], and the DNN, used for speech recognition [28]. In our pilot work [29], we assumed that an important noti cation depends on the noti cation's contents and the user's context.…”