arXiv subjects

He Wen

Publications and source records attributed to He Wen.

28 records · Page 2Linked to original sources

arXiv2020

Simultaneous Sorting of Wavelengths and Spatial Modes using Multi-Plane Light Conversion

We propose a wavelength-mode sorter realized by multi-plane light conversion (MPLC). For the first time, to our best knowledge, wavelengths and spatial modes can be sorted simultaneously. We first demonstrate pure wavelength sorting by a series of phase masks, which could find applications in high-power wavelength beam combining (WBC) or coarse wavelength-division multiplexing (CWDM), for example. We then present a design of a 4-wavelength, 3-mode sorter using only 5 phase masks. Insertion loss (IL) and mode dependent loss (MDL) as low as 1.27 dB and 2.45 dB can be achieved, respectively.

physics.optics↗

arXiv2020

InterHand2.6M: A Dataset and Baseline for 3D Interacting Hand Pose Estimation from a Single RGB Image

Analysis of hand-hand interactions is a crucial step towards better understanding human behavior. However, most researches in 3D hand pose estimation have focused on the isolated single hand case. Therefore, we firstly propose (1) a large-scale dataset, InterHand2.6M, and (2) a baseline network, InterNet, for 3D interacting hand pose estimation from a single RGB image. The proposed InterHand2.6M consists of \textbf{2.6M labeled single and interacting hand frames} under various poses from multiple subjects. Our InterNet simultaneously performs 3D single and interacting hand pose estimation. In our experiments, we demonstrate big gains in 3D interacting hand pose estimation accuracy when leveraging the interacting hand data in InterHand2.6M. We also report the accuracy of InterNet on InterHand2.6M, which serves as a strong baseline for this new dataset. Finally, we show 3D interacting hand pose estimation results from general images. Our code and dataset are available at https://mks0601.github.io/InterHand2.6M/.

cs.CV↗

arXiv2018

DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients

We propose DoReFa-Net, a method to train convolutional neural networks that have low bitwidth weights and activations using low bitwidth parameter gradients. In particular, during backward pass, parameter gradients are stochastically quantized to low bitwidth numbers before being propagated to convolutional layers. As convolutions during forward/backward passes can now operate on low bitwidth weights and activations/gradients respectively, DoReFa-Net can use bit convolution kernels to accelerate both training and inference. Moreover, as bit convolutions can be efficiently implemented on CPU, FPGA, ASIC and GPU, DoReFa-Net opens the way to accelerate training of low bitwidth neural network on these hardware. Our experiments on SVHN and ImageNet datasets prove that DoReFa-Net can achieve comparable prediction accuracy as 32-bit counterparts. For example, a DoReFa-Net derived from AlexNet that has 1-bit weights, 2-bit activations, can be trained from scratch using 6-bit gradients to get 46.1\% top-1 accuracy on ImageNet validation set. The DoReFa-Net AlexNet model is released publicly.

cs.NE↗

arXiv2017

EAST: An Efficient and Accurate Scene Text Detector

Previous approaches for scene text detection have already achieved promising performances across various benchmarks. However, they usually fall short when dealing with challenging scenarios, even when equipped with deep neural network models, because the overall performance is determined by the interplay of multiple stages and components in the pipelines. In this work, we propose a simple yet powerful pipeline that yields fast and accurate text detection in natural scenes. The pipeline directly predicts words or text lines of arbitrary orientations and quadrilateral shapes in full images, eliminating unnecessary intermediate steps (e.g., candidate aggregation and word partitioning), with a single neural network. The simplicity of our pipeline allows concentrating efforts on designing loss functions and neural network architecture. Experiments on standard datasets including ICDAR 2015, COCO-Text and MSRA-TD500 demonstrate that the proposed algorithm significantly outperforms state-of-the-art methods in terms of both accuracy and efficiency. On the ICDAR 2015 dataset, the proposed algorithm achieves an F-score of 0.7820 at 13.2fps at 720p resolution.

cs.CV↗

arXiv2017

Balanced Quantization: An Effective and Efficient Approach to Quantized Neural Networks

Quantized Neural Networks (QNNs), which use low bitwidth numbers for representing parameters and performing computations, have been proposed to reduce the computation complexity, storage size and memory usage. In QNNs, parameters and activations are uniformly quantized, such that the multiplications and additions can be accelerated by bitwise operations. However, distributions of parameters in Neural Networks are often imbalanced, such that the uniform quantization determined from extremal values may under utilize available bitwidth. In this paper, we propose a novel quantization method that can ensure the balance of distributions of quantized values. Our method first recursively partitions the parameters by percentiles into balanced bins, and then applies uniform quantization. We also introduce computationally cheaper approximations of percentiles to reduce the computation overhead introduced. Overall, our method improves the prediction accuracies of QNNs without introducing extra computation during inference, has negligible impact on training speed, and is applicable to both Convolutional Neural Networks and Recurrent Neural Networks. Experiments on standard datasets including ImageNet and Penn Treebank confirm the effectiveness of our method. On ImageNet, the top-5 error rate of our 4-bit quantized GoogLeNet model is 12.7\%, which is superior to the state-of-the-arts of QNNs.

cs.CV↗

arXiv2016

Training Bit Fully Convolutional Network for Fast Semantic Segmentation

Fully convolutional neural networks give accurate, per-pixel prediction for input images and have applications like semantic segmentation. However, a typical FCN usually requires lots of floating point computation and large run-time memory, which effectively limits its usability. We propose a method to train Bit Fully Convolution Network (BFCN), a fully convolutional neural network that has low bit-width weights and activations. Because most of its computation-intensive convolutions are accomplished between low bit-width numbers, a BFCN can be accelerated by an efficient bit-convolution implementation. On CPU, the dot product operation between two bit vectors can be reduced to bitwise operations and popcounts, which can offer much higher throughput than 32-bit multiplications and additions. To validate the effectiveness of BFCN, we conduct experiments on the PASCAL VOC 2012 semantic segmentation task and Cityscapes. Our BFCN with 1-bit weights and 2-bit activations, which runs 7.8x faster on CPU or requires less than 1\% resources on FPGA, can achieve comparable performance as the 32-bit counterpart.

cs.CV↗

arXiv2016

Effective Quantization Methods for Recurrent Neural Networks

Reducing bit-widths of weights, activations, and gradients of a Neural Network can shrink its storage size and memory usage, and also allow for faster training and inference by exploiting bitwise operations. However, previous attempts for quantization of RNNs show considerable performance degradation when using low bit-width weights and activations. In this paper, we propose methods to quantize the structure of gates and interlinks in LSTM and GRU cells. In addition, we propose balanced quantization methods for weights to further reduce performance degradation. Experiments on PTB and IMDB datasets confirm effectiveness of our methods as performances of our models match or surpass the previous state-of-the-art of quantized RNN.

cs.LG↗

arXiv2014

Facts, myths and fights about the KLJN classical physical key exchanger

This paper deals with the Kirchhoff-law-Johnson-noise (KLJN) classical statistical physical key exchange method and surveys criticism - often stemming from a lack of understanding of its underlying premises or from other errors - and our related responses against these, often unphysical, claims. Some of the attacks are valid, however, an extended KLJN system remains protected against all of them, implying that its unconditional security is not impacted.

cs.ET↗

arXiv2014

Bird's-eye view on Noise-Based Logic

Noise-based logic is a practically deterministic logic scheme inspired by the randomness of neural spikes and uses a system of uncorrelated stochastic processes and their superposition to represent the logic state. We briefly discuss various questions such as (i) What does practical determinism mean? (ii) Is noise-based logic a Turing machine? (iii) Is there hope to beat (the dreams of) quantum computation by a classical physical noise-based processor, and what are the minimum hardware requirements for that? Finally, (iv) we address the problem of random number generators and show that the common belief that quantum number generators are superior to classical (thermal) noise-based generators is nothing but a myth.

cs.ET↗

arXiv2012

Noise based logic: why noise? A comparative study of the necessity of randomness out of orthogonality

Although noise-based logic shows potential advantages of reduced power dissipation and the ability of large parallel operations with low hardware and time complexity the question still persist: is randomness really needed out of orthogonality? In this Letter, after some general thermodynamical considerations, we show relevant examples where we compare the computational complexity of logic systems based on orthogonal noise and sinusoidal signals, respectively. The conclusion is that in certain special-purpose applications noise-based logic is exponentially better than its sinusoidal version: its computational complexity can be exponentially smaller to perform the same task.

cs.OH↗