arXiv subjects

Kwang-Ting Cheng

Publications and source records attributed to Kwang-Ting Cheng.

108 records · Page 6Linked to original sources

arXiv2021

Joint Multi-Dimension Pruning via Numerical Gradient Update

We present joint multi-dimension pruning (abbreviated as JointPruning), an effective method of pruning a network on three crucial aspects: spatial, depth and channel simultaneously. To tackle these three naturally different dimensions, we proposed a general framework by defining pruning as seeking the best pruning vector (i.e., the numerical value of layer-wise channel number, spacial size, depth) and construct a unique mapping from the pruning vector to the pruned network structures. Then we optimize the pruning vector with gradient update and model joint pruning as a numerical gradient optimization process. To overcome the challenge that there is no explicit function between the loss and the pruning vectors, we proposed self-adapted stochastic gradient estimation to construct a gradient path through network loss to pruning vectors and enable efficient gradient update. We show that the joint strategy discovers a better status than previous studies that focused on individual dimensions solely, as our method is optimized collaboratively across the three dimensions in a single end-to-end training and it is more efficient than the previous exhaustive methods. Extensive experiments on large-scale ImageNet dataset across a variety of network architectures MobileNet V1&V2&V3 and ResNet demonstrate the effectiveness of our proposed method. For instance, we achieve significant margins of 2.5% and 2.6% improvement over the state-of-the-art approach on the already compact MobileNet V1&V2 under an extremely large compression ratio.

cs.CV↗

arXiv2021

Exploring intermediate representation for monocular vehicle pose estimation

We present a new learning-based framework to recover vehicle pose in SO(3) from a single RGB image. In contrast to previous works that map from local appearance to observation angles, we explore a progressive approach by extracting meaningful Intermediate Geometrical Representations (IGRs) to estimate egocentric vehicle orientation. This approach features a deep model that transforms perceived intensities to IGRs, which are mapped to a 3D representation encoding object orientation in the camera coordinate system. Core problems are what IGRs to use and how to learn them more effectively. We answer the former question by designing IGRs based on an interpolated cuboid that derives from primitive 3D annotation readily. The latter question motivates us to incorporate geometry knowledge with a new loss function based on a projective invariant. This loss function allows unlabeled data to be used in the training stage to improve representation learning. Without additional labels, our system outperforms previous monocular RGB-based methods for joint vehicle detection and pose estimation on the KITTI benchmark, achieving performance even comparable to stereo methods. Code and pre-trained models are available at this https URL.

cs.CV↗

arXiv2021

R2F: A Remote Retraining Framework for AIoT Processors with Computing Errors

AIoT processors fabricated with newer technology nodes suffer rising soft errors due to the shrinking transistor sizes and lower power supply. Soft errors on the AIoT processors particularly the deep learning accelerators (DLAs) with massive computing may cause substantial computing errors. These computing errors are difficult to be captured by the conventional training on general purposed processors like CPUs and GPUs in a server. Applying the offline trained neural network models to the edge accelerators with errors directly may lead to considerable prediction accuracy loss. To address the problem, we propose a remote retraining framework (R2F) for remote AIoT processors with computing errors. It takes the remote AIoT processor with soft errors in the training loop such that the on-site computing errors can be learned with the application data on the server and the retrained models can be resilient to the soft errors. Meanwhile, we propose an optimized partial TMR strategy to enhance the retraining. According to our experiments, R2F enables elastic design trade-offs between the model accuracy and the performance penalty. The top-5 model accuracy can be improved by 1.93%-13.73% with 0%-200% performance penalty at high fault error rate. In addition, we notice that the retraining requires massive data transmission and even dominates the training time, and propose a sparse increment compression approach for the data transmission optimization, which reduces the retraining time by 38%-88% on average with negligible accuracy loss over a straightforward remote retraining.

cs.AR↗

arXiv2021

Energy-Efficient Accelerator Design for Deformable Convolution Networks

Deformable convolution networks (DCNs) proposed to address the image recognition with geometric or photometric variations typically involve deformable convolution that convolves on arbitrary locations of input features. The locations change with different inputs and induce considerable dynamic and irregular memory accesses which cannot be handled by classic neural network accelerators (NNAs). Moreover, bilinear interpolation (BLI) operation that is required to obtain deformed features in DCNs also cannot be deployed on existing NNAs directly. Although a general purposed processor (GPP) seated along with classic NNAs can process the deformable convolution, the processing on GPP can be extremely slow due to the lack of parallel computing capability. To address the problem, we develop a DCN accelerator on existing NNAs to support both the standard convolution and deformable convolution. Specifically, for the dynamic and irregular accesses in DCNs, we have both the input and output features divided into tiles and build a tile dependency table (TDT) to track the irregular tile dependency at runtime. With the TDT, we further develop an on-chip tile scheduler to handle the dynamic and irregular accesses efficiently. In addition, we propose a novel mapping strategy to enable parallel BLI processing on NNAs and apply layer fusion techniques for more energy-efficient DCN processing. According to our experiments, the proposed accelerator achieves orders of magnitude higher performance and energy efficiency compared to the typical computing architectures including ARM, ARM+TPU, and GPU with 6.6\% chip area penalty to a classic NNA.

cs.AR↗

arXiv2021

S2-BNN: Bridging the Gap Between Self-Supervised Real and 1-bit Neural Networks via Guided Distribution Calibration

Previous studies dominantly target at self-supervised learning on real-valued networks and have achieved many promising results. However, on the more challenging binary neural networks (BNNs), this task has not yet been fully explored in the community. In this paper, we focus on this more difficult scenario: learning networks where both weights and activations are binary, meanwhile, without any human annotated labels. We observe that the commonly used contrastive objective is not satisfying on BNNs for competitive accuracy, since the backbone network contains relatively limited capacity and representation ability. Hence instead of directly applying existing self-supervised methods, which cause a severe decline in performance, we present a novel guided learning paradigm from real-valued to distill binary networks on the final prediction distribution, to minimize the loss and obtain desirable accuracy. Our proposed method can boost the simple contrastive learning baseline by an absolute gain of 5.5~15% on BNNs. We further reveal that it is difficult for BNNs to recover the similar predictive distributions as real-valued models when training without labels. Thus, how to calibrate them is key to address the degradation in performance. Extensive experiments are conducted on the large-scale ImageNet and downstream datasets. Our method achieves substantial improvement over the simple contrastive learning baseline, and is even comparable to many mainstream supervised BNN methods. Code is available at https://github.com/szq0214/S2-BNN.

cs.CV↗

arXiv2021

How Do Adam and Training Strategies Help BNNs Optimization?

The best performing Binary Neural Networks (BNNs) are usually attained using Adam optimization and its multi-step training variants. However, to the best of our knowledge, few studies explore the fundamental reasons why Adam is superior to other optimizers like SGD for BNN optimization or provide analytical explanations that support specific training strategies. To address this, in this paper we first investigate the trajectories of gradients and weights in BNNs during the training process. We show the regularization effect of second-order momentum in Adam is crucial to revitalize the weights that are dead due to the activation saturation in BNNs. We find that Adam, through its adaptive learning rate strategy, is better equipped to handle the rugged loss surface of BNNs and reaches a better optimum with higher generalization ability. Furthermore, we inspect the intriguing role of the real-valued weights in binary networks, and reveal the effect of weight decay on the stability and sluggishness of BNN optimization. Through extensive experiments and analysis, we derive a simple training scheme, building on existing Adam-based optimization, which achieves 70.5% top-1 accuracy on the ImageNet dataset using the same architecture as the state-of-the-art ReActNet while achieving 1.1% higher accuracy. Code and models are available at https://github.com/liuzechun/AdamBNN.

cs.LG↗

arXiv2021

Cascaded deep monocular 3D human pose estimation with evolutionary training data

End-to-end deep representation learning has achieved remarkable accuracy for monocular 3D human pose estimation, yet these models may fail for unseen poses with limited and fixed training data. This paper proposes a novel data augmentation method that: (1) is scalable for synthesizing massive amount of training data (over 8 million valid 3D human poses with corresponding 2D projections) for training 2D-to-3D networks, (2) can effectively reduce dataset bias. Our method evolves a limited dataset to synthesize unseen 3D human skeletons based on a hierarchical human representation and heuristics inspired by prior knowledge. Extensive experiments show that our approach not only achieves state-of-the-art accuracy on the largest public benchmark, but also generalizes significantly better to unseen and rare poses. Code, pre-trained models and tools are available at this HTTPS URL.

cs.CV↗

arXiv2021

Is Label Smoothing Truly Incompatible with Knowledge Distillation: An Empirical Study

This work aims to empirically clarify a recently discovered perspective that label smoothing is incompatible with knowledge distillation. We begin by introducing the motivation behind on how this incompatibility is raised, i.e., label smoothing erases relative information between teacher logits. We provide a novel connection on how label smoothing affects distributions of semantically similar and dissimilar classes. Then we propose a metric to quantitatively measure the degree of erased information in sample's representation. After that, we study its one-sidedness and imperfection of the incompatibility view through massive analyses, visualizations and comprehensive experiments on Image Classification, Binary Networks, and Neural Machine Translation. Finally, we broadly discuss several circumstances wherein label smoothing will indeed lose its effectiveness. Project page: http://zhiqiangshen.com/projects/LS_and_KD/index.html.

cs.LG↗

arXiv2021

Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning

The goal of few-shot learning is to learn a classifier that can recognize unseen classes from limited support data with labels. A common practice for this task is to train a model on the base set first and then transfer to novel classes through fine-tuning (Here fine-tuning procedure is defined as transferring knowledge from base to novel data, i.e. learning to transfer in few-shot scenario.) or meta-learning. However, as the base classes have no overlap to the novel set, simply transferring whole knowledge from base data is not an optimal solution since some knowledge in the base model may be biased or even harmful to the novel class. In this paper, we propose to transfer partial knowledge by freezing or fine-tuning particular layer(s) in the base model. Specifically, layers will be imposed different learning rates if they are chosen to be fine-tuned, to control the extent of preserved transferability. To determine which layers to be recast and what values of learning rates for them, we introduce an evolutionary search based method that is efficient to simultaneously locate the target layers and determine their individual learning rates. We conduct extensive experiments on CUB and mini-ImageNet to demonstrate the effectiveness of our proposed method. It achieves the state-of-the-art performance on both meta-learning and non-meta based frameworks. Furthermore, we extend our method to the conventional pre-training + fine-tuning paradigm and obtain consistent improvement.

cs.CV↗

arXiv2020

ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions

In this paper, 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, bypassing all the intermediate convolutional layers including the downsampling layers. This baseline network strikes a good trade-off between accuracy and efficiency, achieving superior performance than most of existing binary networks at approximately half of the computational cost. Through extensive experiments and analysis, we observed that the performance of binary networks is sensitive to activation distribution variations. Based on this important observation, we propose to generalize the traditional Sign and PReLU functions, denoted as RSign and RPReLU for the respective generalized functions, to enable explicit learning of the distribution reshape and shift at near-zero extra cost. Lastly, we adopt a distributional loss to further enforce the binary network to learn similar output distributions as those of a real-valued network. We show that after incorporating all these ideas, the proposed ReActNet outperforms all the state-of-the-arts by a large margin. Specifically, it outperforms Real-to-Binary Net and MeliusNet29 by 4.0% and 3.6% respectively for the top-1 accuracy and also reduces the gap to its real-valued counterpart to within 3.0% top-1 accuracy on ImageNet dataset. Code and models are available at: https://github.com/liuzechun/ReActNet.

cs.CV↗

arXiv2020

Binarizing MobileNet via Evolution-based Searching

Binary Neural Networks (BNNs), known to be one among the effectively compact network architectures, have achieved great outcomes in the visual tasks. Designing efficient binary architectures is not trivial due to the binary nature of the network. In this paper, we propose a use of evolutionary search to facilitate the construction and training scheme when binarizing MobileNet, a compact network with separable depth-wise convolution. Inspired by one-shot architecture search frameworks, we manipulate the idea of group convolution to design efficient 1-Bit Convolutional Neural Networks (CNNs), assuming an approximately optimal trade-off between computational cost and model accuracy. Our objective is to come up with a tiny yet efficient binary neural architecture by exploring the best candidates of the group convolution while optimizing the model performance in terms of complexity and latency. The approach is threefold. First, we train strong baseline binary networks with a wide range of random group combinations at each convolutional layer. This set-up gives the binary neural networks a capability of preserving essential information through layers. Second, to find a good set of hyperparameters for group convolutions we make use of the evolutionary search which leverages the exploration of efficient 1-bit models. Lastly, these binary models are trained from scratch in a usual manner to achieve the final binary model. Various experiments on ImageNet are conducted to show that following our construction guideline, the final model achieves 60.09% Top-1 accuracy and outperforms the state-of-the-art CI-BCNN with the same computational cost.

cs.CV↗

arXiv2019

Label Refinement with an Iterative Generative Adversarial Network for Boosting Retinal Vessel Segmentation

State-of-the-art methods for retinal vessel segmentation mainly rely on manually labeled vessels as the ground truth for supervised training. The quality of manual labels plays an essential role in the segmentation accuracy, while in practice it could vary a lot and in turn could substantially mislead the training process and limit the segmentation accuracy. This paper aims to "purify" any comprehensive training set, which consists of data annotated by various observers, via refining low-quality manual labels in the dataset. To this end, we have developed a novel label refinement method based on an iterative generative adversarial network (GAN). Our iterative GAN is trained based on a set of high-quality patches (i.e. with consistent manual labels among different observers) and low-quality patches with noisy manual vessel labels. A simple yet effective method has been designed to simulate low-quality patches with noises which conform to the distribution of real noises from human observers. To evaluate the effectiveness of our method, we have trained four state-of-the-art retinal vessel segmentation models using the purified dataset obtained from our method and compared their performance with those trained based on the original noisy datasets. Experimental results on two datasets DRIVE and CHASE_DB1 demonstrate that obvious accuracy improvements can be achieved for all the four models when using the purified datasets from our method.

eess.IV↗

arXiv2019

Latent Weights Do Not Exist: Rethinking Binarized Neural Network Optimization

Optimization of Binarized Neural Networks (BNNs) currently relies on real-valued latent weights to accumulate small update steps. In this paper, we argue that these latent weights cannot be treated analogously to weights in real-valued networks. Instead their main role is to provide inertia during training. We interpret current methods in terms of inertia and provide novel insights into the optimization of BNNs. We subsequently introduce the first optimizer specifically designed for BNNs, Binary Optimizer (Bop), and demonstrate its performance on CIFAR-10 and ImageNet. Together, the redefinition of latent weights as inertia and the introduction of Bop enable a better understanding of BNN optimization and open up the way for further improvements in training methodologies for BNNs. Code is available at: https://github.com/plumerai/rethinking-bnn-optimization

cs.LG↗

arXiv2019

Bi-Real Net: Binarizing Deep Network Towards Real-Network Performance

In this paper, we study 1-bit convolutional neural networks (CNNs), of which both the weights and activations are binary. While efficient, the lacking of representational capability and the training difficulty impede 1-bit CNNs from performing as well as real-valued networks. We propose Bi-Real net with a novel training algorithm to tackle these two challenges. To enhance the representational capability, we propagate the real-valued activations generated by each 1-bit convolution via a parameter-free shortcut. To address the training difficulty, we propose a training algorithm using a tighter approximation to the derivative of the sign function, a magnitude-aware gradient for weight updating, a better initialization method, and a two-step scheme for training a deep network. Experiments on ImageNet show that an 18-layer Bi-Real net with the proposed training algorithm achieves 56.4% top-1 classification accuracy, which is 10% higher than the state-of-the-arts (e.g., XNOR-Net) with greater memory saving and lower computational cost. Bi-Real net is also the first to scale up 1-bit CNNs to an ultra-deep network with 152 layers, and achieves 64.5% top-1 accuracy on ImageNet. A 50-layer Bi-Real net shows comparable performance to a real-valued network on the depth estimation task with only a 0.3% accuracy gap.

cs.CV↗

arXiv2019

Facial age estimation by deep residual decision making

Residual representation learning simplifies the optimization problem of learning complex functions and has been widely used by traditional convolutional neural networks. However, it has not been applied to deep neural decision forest (NDF). In this paper we incorporate residual learning into NDF and the resulting model achieves state-of-the-art level accuracy on three public age estimation benchmarks while requiring less memory and computation. We further employ gradient-based technique to visualize the decision-making process of NDF and understand how it is influenced by facial image inputs. The code and pre-trained models will be available at https://github.com/Nicholasli1995/VisualizingNDF.

cs.CV↗

arXiv2019

Visualizing the decision-making process in deep neural decision forest

Deep neural decision forest (NDF) achieved remarkable performance on various vision tasks via combining decision tree and deep representation learning. In this work, we first trace the decision-making process of this model and visualize saliency maps to understand which portion of the input influence it more for both classification and regression problems. We then apply NDF on a multi-task coordinate regression problem and demonstrate the distribution of routing probabilities, which is vital for interpreting NDF yet not shown for regression problems. The pre-trained model and code for visualization will be available at https://github.com/Nicholasli1995/VisualizingNDF

cs.CV↗

arXiv2018

Semi-supervised mp-MRI Data Synthesis with StitchLayer and Auxiliary Distance Maximization

In this paper, we address the problem of synthesizing multi-parameter magnetic resonance imaging (mp-MRI) data, i.e. Apparent Diffusion Coefficients (ADC) and T2-weighted (T2w), containing clinically significant (CS) prostate cancer (PCa) via semi-supervised adversarial learning. Specifically, our synthesizer generates mp-MRI data in a sequential manner: first generating ADC maps from 128-d latent vectors, followed by translating them to the T2w images. The synthesizer is trained in a semisupervised manner. In the supervised training process, a limited amount of paired ADC-T2w images and the corresponding ADC encodings are provided and the synthesizer learns the paired relationship by explicitly minimizing the reconstruction losses between synthetic and real images. To avoid overfitting limited ADC encodings, an unlimited amount of random latent vectors and unpaired ADC-T2w Images are utilized in the unsupervised training process for learning the marginal image distributions of real images. To improve the robustness of synthesizing, we decompose the difficult task of generating full-size images into several simpler tasks which generate sub-images only. A StitchLayer is then employed to fuse sub-images together in an interlaced manner into a full-size image. To enforce the synthetic images to indeed contain distinguishable CS PCa lesions, we propose to also maximize an auxiliary distance of Jensen-Shannon divergence (JSD) between CS and nonCS images. Experimental results show that our method can effectively synthesize a large variety of mpMRI images which contain meaningful CS PCa lesions, display a good visual quality and have the correct paired relationship. Compared to the state-of-the-art synthesis methods, our method achieves a significant improvement in terms of both visual and quantitative evaluation metrics.

cs.LG↗

arXiv2018

Bi-Real Net: Enhancing the Performance of 1-bit CNNs With Improved Representational Capability and Advanced Training Algorithm

In this work, we study the 1-bit convolutional neural networks (CNNs), of which both the weights and activations are binary. While being efficient, the classification accuracy of the current 1-bit CNNs is much worse compared to their counterpart real-valued CNN models on the large-scale dataset, like ImageNet. To minimize the performance gap between the 1-bit and real-valued CNN models, we propose a novel model, dubbed Bi-Real net, which connects the real activations (after the 1-bit convolution and/or BatchNorm layer, before the sign function) to activations of the consecutive block, through an identity shortcut. Consequently, compared to the standard 1-bit CNN, the representational capability of the Bi-Real net is significantly enhanced and the additional cost on computation is negligible. Moreover, we develop a specific training algorithm including three technical novelties for 1- bit CNNs. Firstly, we derive a tight approximation to the derivative of the non-differentiable sign function with respect to activation. Secondly, we propose a magnitude-aware gradient with respect to the weight for updating the weight parameters. Thirdly, we pre-train the real-valued CNN model with a clip function, rather than the ReLU function, to better initialize the Bi-Real net. Experiments on ImageNet show that the Bi-Real net with the proposed training algorithm achieves 56.4% and 62.2% top-1 accuracy with 18 layers and 34 layers, respectively. Compared to the state-of-the-arts (e.g., XNOR Net), Bi-Real net achieves up to 10% higher top-1 accuracy with more memory saving and lower computational cost. Keywords: binary neural network, 1-bit CNNs, 1-layer-per-block

cs.CV↗