PyTorchKR Papers

최신 논문

총 2,153편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2208.03918공식 코드 · Papers with Code

Depth Quality-Inspired Feature Manipulation for Efficient RGB-D and Video Salient Object Detection

Recently CNN-based RGB-D salient object detection (SOD) has obtained significant improvement on detection accuracy. However, existing models often fail to perform well in terms of efficiency and accuracy simultaneously. This hinders their potential applications on mobile devices as well as many real-world problems. To bridge the accuracy gap between lightweight and large models for RGB-D SOD, in this paper, an efficient module that can greatly improve the accuracy but adds little computation is proposed. Inspired by the fact that depth quality is a key factor influencing the accuracy, we propose an efficient depth quality-inspired feature manipulation (DQFM) process, which can dynamically filter depth features according to depth quality. The proposed DQFM resorts to the alignment of low-level RGB and depth features, as well as holistic attention of the depth stream to explicitly control and enhance cross-modal fusion. We embed DQFM to obtain an efficient lightweight RGB-D SOD model called DFM-Net, where we in addition design a tailored depth backbone and a two-stage decoder as basic parts. Extensive experimental results on nine RGB-D datasets demonstrate that our DFM-Net outperforms recent efficient models, running at about 20 FPS on CPU with only 8.5Mb model size, and meanwhile being 2.9/2.4 times faster and 6.7/3.1 times smaller than the latest best models A2dele and MobileSal. It also maintains state-of-the-art accuracy when even compared to non-efficient models. Interestingly, further statistics and analyses verify the ability of DQFM in distinguishing depth maps of various qualities without any quality labels. Last but not least, we further apply DFM-Net to deal with video SOD (VSOD), achieving comparable performance against recent efficient models while being 3/2.3 times faster/smaller than the prior best in this field. Our code is available at https://github.com/zwbx/DFM-Net.

arXiv:2208.01641공식 코드 · Papers with Code

Streaming-capable High-performance Architecture of Learned Image Compression Codecs

Learned image compression allows achieving state-of-the-art accuracy and compression ratios, but their relatively slow runtime performance limits their usage. While previous attempts on optimizing learned image codecs focused more on the neural model and entropy coding, we present an alternative method to improving the runtime performance of various learned image compression models. We introduce multi-threaded pipelining and an optimized memory model to enable GPU and CPU workloads asynchronous execution, fully taking advantage of computational resources. Our architecture alone already produces excellent performance without any change to the neural model itself. We also demonstrate that combining our architecture with previous tweaks to the neural models can further improve runtime performance. We show that our implementations excel in throughput and latency compared to the baseline and demonstrate the performance of our implementations by creating a real-time video streaming encoder-decoder sample application, with the encoder running on an embedded device.

arXiv:2208.00763

HiKonv: Maximizing the Throughput of Quantized Convolution With Novel Bit-wise Management and Computation

Quantization for CNN has shown significant progress with the intention of reducing the cost of computation and storage with low-bitwidth data representations. There are, however, no systematic studies on how an existing full-bitwidth processing unit, such as ALU in CPUs and DSP in FPGAs, can be better utilized to deliver significantly higher computation throughput for convolution under various quantized bitwidths. In this study, we propose HiKonv, a unified solution that maximizes the throughput of convolution on a given underlying processing unit with low-bitwidth quantized data inputs through novel bit-wise management and parallel computation. We establish theoretical framework and performance models using a full-bitwidth multiplier for highly parallelized low-bitwidth convolution, and demonstrate new breakthroughs for high-performance computing in this critical domain. For example, a single 32-bit processing unit in CPU can deliver 128 binarized convolution operations (multiplications and additions) and 13 4-bit convolution operations with a single multiplication instruction, and a single 27x18 multiplier in the FPGA DSP can deliver 60, 8 or 2 convolution operations with 1, 4 or 8-bit inputs in one clock cycle. We demonstrate the effectiveness of HiKonv on both CPU and FPGA. On CPU, HiKonv outperforms the baseline implementation with 1 to 8-bit inputs and provides up to 7.6x and 1.4x performance improvements for 1-D convolution, and performs 2.74x and 3.19x over the baseline implementation for 4-bit signed and unsigned data inputs for 2-D convolution. On FPGA, HiKonv solution enables a single DSP to process multiple convolutions with a shorter processing latency. For binarized input, each DSP with HiKonv is equivalent up to 76.6 LUTs. Compared to the DAC-SDC 2020 champion model, HiKonv achieves a 2.37x throughput improvement and 2.61x DSP efficiency improvement, respectively.

arXiv:2208.00001공식 코드 · Papers with Code

FastGeodis: Fast Generalised Geodesic Distance Transform

The FastGeodis package provides an efficient implementation for computing Geodesic and Euclidean distance transforms (or a mixture of both), targeting efficient utilisation of CPU and GPU hardware. In particular, it implements the paralellisable raster scan method from Criminisi et al. (2009), where elements in a row (2D) or plane (3D) can be computed with parallel threads. This package is able to handle 2D as well as 3D data, where it achieves up to a 20x speedup on a CPU and up to a 74x speedup on a GPU as compared to an existing open-source library (Wang, 2020) that uses a non-parallelisable single-thread CPU implementation. The performance speedups reported here were evaluated using 3D volume data on an Nvidia GeForce Titan X (12 GB) with a 6-Core Intel Xeon E5-1650 CPU. Further in-depth comparison of performance improvements are discussed in the FastGeodis documentation: https://fastgeodis.readthedocs.io

arXiv:2208.00800

GANDSE: Generative Adversarial Network based Design Space Exploration for Neural Network Accelerator Design

With the popularity of deep learning, the hardware implementation platform of deep learning has received increasing interest. Unlike the general purpose devices, e.g., CPU, or GPU, where the deep learning algorithms are executed at the software level, neural network hardware accelerators directly execute the algorithms to achieve higher both energy efficiency and performance improvements. However, as the deep learning algorithms evolve frequently, the engineering effort and cost of designing the hardware accelerators are greatly increased. To improve the design quality while saving the cost, design automation for neural network accelerators was proposed, where design space exploration algorithms are used to automatically search the optimized accelerator design within a design space. Nevertheless, the increasing complexity of the neural network accelerators brings the increasing dimensions to the design space. As a result, the previous design space exploration algorithms are no longer effective enough to find an optimized design. In this work, we propose a neural network accelerator design automation framework named GANDSE, where we rethink the problem of design space exploration, and propose a novel approach based on the generative adversarial network (GAN) to support an optimized exploration for high dimension large design space. The experiments show that GANDSE is able to find the more optimized designs in negligible time compared with approaches including multilayer perceptron and deep reinforcement learning.

arXiv:2208.00498

DNNShield: Dynamic Randomized Model Sparsification, A Defense Against Adversarial Machine Learning

DNNs are known to be vulnerable to so-called adversarial attacks that manipulate inputs to cause incorrect results that can be beneficial to an attacker or damaging to the victim. Recent works have proposed approximate computation as a defense mechanism against machine learning attacks. We show that these approaches, while successful for a range of inputs, are insufficient to address stronger, high-confidence adversarial attacks. To address this, we propose DNNSHIELD, a hardware-accelerated defense that adapts the strength of the response to the confidence of the adversarial input. Our approach relies on dynamic and random sparsification of the DNN model to achieve inference approximation efficiently and with fine-grain control over the approximation error. DNNSHIELD uses the output distribution characteristics of sparsified inference compared to a dense reference to detect adversarial inputs. We show an adversarial detection rate of 86% when applied to VGG16 and 88% when applied to ResNet50, which exceeds the detection rate of the state of the art approaches, with a much lower overhead. We demonstrate a software/hardware-accelerated FPGA prototype, which reduces the performance impact of DNNSHIELD relative to software-only CPU and GPU implementations.

arXiv:2207.14524

Evaluating the Practicality of Learned Image Compression

Learned image compression has achieved extraordinary rate-distortion performance in PSNR and MS-SSIM compared to traditional methods. However, it suffers from intensive computation, which is intolerable for real-world applications and leads to its limited industrial application for now. In this paper, we introduce neural architecture search (NAS) to designing more efficient networks with lower latency, and leverage quantization to accelerate the inference process. Meanwhile, efforts in engineering like multi-threading and SIMD have been made to improve efficiency. Optimized using a hybrid loss of PSNR and MS-SSIM for better visual quality, we obtain much higher MS-SSIM than JPEG, JPEG XL and AVIF over all bit rates, and PSNR between that of JPEG XL and AVIF. Our software implementation of LIC achieves comparable or even faster inference speed compared to jpeg-turbo while being multiple times faster than JPEG XL and AVIF. Besides, our implementation of LIC reaches stunning throughput of 145 fps for encoding and 208 fps for decoding on a Tesla T4 GPU for 1080p images. On CPU, the latency of our implementation is comparable with JPEG XL.

arXiv:2207.12547공식 코드 · Papers with Code

The BUTTER Zone: An Empirical Study of Training Dynamics in Fully Connected Neural Networks

We present an empirical dataset surveying the deep learning phenomenon on fully-connected feed-forward multilayer perceptron neural networks. The dataset, which is now freely available online, records the per-epoch training and generalization performance of 483 thousand distinct hyperparameter choices of architectures, tasks, depths, network sizes (number of parameters), learning rates, batch sizes, and regularization penalties. Repeating each experiment an average of 24 times resulted in 11 million total training runs and 40 billion epochs recorded. Accumulating this 1.7 TB dataset utilized 11 thousand CPU core-years, 72.3 GPU-years, and 163 node-years. In surveying the dataset, we observe durable patterns persisting across tasks and topologies. We aim to spark scientific study of machine learning techniques as a catalyst for the theoretical discoveries needed to progress the field beyond energy-intensive and heuristic practices.

arXiv:2207.12620공식 코드 · Papers with Code

Large-displacement 3D Object Tracking with Hybrid Non-local Optimization

Optimization-based 3D object tracking is known to be precise and fast, but sensitive to large inter-frame displacements. In this paper we propose a fast and effective non-local 3D tracking method. Based on the observation that erroneous local minimum are mostly due to the out-of-plane rotation, we propose a hybrid approach combining non-local and local optimizations for different parameters, resulting in efficient non-local search in the 6D pose space. In addition, a precomputed robust contour-based tracking method is proposed for the pose optimization. By using long search lines with multiple candidate correspondences, it can adapt to different frame displacements without the need of coarse-to-fine search. After the pre-computation, pose updates can be conducted very fast, enabling the non-local optimization to run in real time. Our method outperforms all previous methods for both small and large displacements. For large displacements, the accuracy is greatly improved ($81.7\% \;\text{v.s.}\; 19.4\%$). At the same time, real-time speed ($>$50fps) can be achieved with only CPU. The source code is available at \url{https://github.com/cvbubbles/nonlocal-3dtracking}.

arXiv:2207.11312

HybMT: Hybrid Meta-Predictor based ML Algorithm for Fast Test Vector Generation

ML models are increasingly being used to increase the test coverage and decrease the overall testing time. This field is still in its nascent stage and up till now there were no algorithms that could match or outperform commercial tools in terms of speed and accuracy for large circuits. We propose an ATPG algorithm HybMT in this paper that finally breaks this barrier. Like sister methods, we augment the classical PODEM algorithm that uses recursive backtracking. We design a custom 2-level predictor that predicts the input net of a logic gate whose value needs to be set to ensure that the output is a given value (0 or 1). Our predictor chooses the output from among two first-level predictors, where the most effective one is a bespoke neural network and the other is an SVM regressor. As compared to a popular, state-of-the-art commercial ATPG tool, HybMT shows an overall reduction of 56.6% in the CPU time without compromising on the fault coverage for the EPFL benchmark circuits. HybMT also shows a speedup of 126.4% over the best ML-based algorithm while obtaining an equal or better fault coverage for the EPFL benchmark circuits.

arXiv:2207.11512공식 코드 · Papers with Code

Combining Self-Training and Hybrid Architecture for Semi-supervised Abdominal Organ Segmentation

Abdominal organ segmentation has many important clinical applications, such as organ quantification, surgical planning, and disease diagnosis. However, manually annotating organs from CT scans is time-consuming and labor-intensive. Semi-supervised learning has shown the potential to alleviate this challenge by learning from a large set of unlabeled images and limited labeled samples. In this work, we follow the self-training strategy and employ a high-performance hybrid architecture (PHTrans) consisting of CNN and Swin Transformer for the teacher model to generate precise pseudo labels for unlabeled data. Afterward, we introduce them with labeled data together into a two-stage segmentation framework with lightweight PHTrans for training to improve the performance and generalization ability of the model while remaining efficient. Experiments on the validation set of FLARE2022 demonstrate that our method achieves excellent segmentation performance as well as fast and low-resource model inference. The average DSC and NSD are 0.8956 and 0.9316, respectively. Under our development environments, the average inference time is 18.62 s, the average maximum GPU memory is 1995.04 MB, and the area under the GPU memory-time curve and the average area under the CPU utilization-time curve are 23196.84 and 319.67. The code is available at https://github.com/lseventeen/FLARE22-TwoStagePHTrans.

arXiv:2207.12116

A Variant of Concurrent Constraint Programming on GPU

The number of cores on graphical computing units (GPUs) is reaching thousands nowadays, whereas the clock speed of processors stagnates. Unfortunately, constraint programming solvers do not take advantage yet of GPU parallelism. One reason is that constraint solvers were primarily designed within the mental frame of sequential computation. To solve this issue, we take a step back and contribute to a simple, intrinsically parallel, lock-free and formally correct programming language based on concurrent constraint programming. We then re-examine parallel constraint solving on GPUs within this formalism, and develop Turbo, a simple constraint solver entirely programmed on GPUs. Turbo validates the correctness of our approach and compares positively to a parallel CPU-based solver.

arXiv:2207.10482공식 코드 · Papers with Code

LPYOLO: Low Precision YOLO for Face Detection on FPGA

In recent years, number of edge computing devices and artificial intelligence applications on them have advanced excessively. In edge computing, decision making processes and computations are moved from servers to edge devices. Hence, cheap and low power devices are required. FPGAs are very low power, inclined to do parallel operations and deeply suitable devices for running Convolutional Neural Networks (CNN) which are the fundamental unit of an artificial intelligence application. Face detection on surveillance systems is the most expected application on the security market. In this work, TinyYolov3 architecture is redesigned and deployed for face detection. It is a CNN based object detection method and developed for embedded systems. PYNQ-Z2 is selected as a target board which has low-end Xilinx Zynq 7020 System-on-Chip (SoC) on it. Redesigned TinyYolov3 model is defined in numerous bit width precisions with Brevitas library which brings fundamental CNN layers and activations in integer quantized form. Then, the model is trained in a quantized structure with WiderFace dataset. In order to decrease latency and power consumption, onchip memory of the FPGA is configured as a storage of whole network parameters and the last activation function is modified as rescaled HardTanh instead of Sigmoid. Also, high degree of parallelism is applied to logical resources of the FPGA. The model is converted to an HLS based application with using FINN framework and FINN-HLS library which includes the layer definitions in C++. Later, the model is synthesized and deployed. CPU of the SoC is employed with multithreading mechanism and responsible for preprocessing, postprocessing and TCP/IP streaming operations. Consequently, 2.4 Watt total board power consumption, 18 Frames-Per-Second (FPS) throughput and 0.757 mAP accuracy rate on Easy category of the WiderFace are achieved with 4 bits precision model.

arXiv:2207.10446

COBRA: Cpu-Only aBdominal oRgan segmentAtion

Abdominal organ segmentation is a difficult and time-consuming task. To reduce the burden on clinical experts, fully-automated methods are highly desirable. Current approaches are dominated by Convolutional Neural Networks (CNNs) however the computational requirements and the need for large data sets limit their application in practice. By implementing a small and efficient custom 3D CNN, compiling the trained model and optimizing the computational graph: our approach produces high accuracy segmentations (Dice Similarity Coefficient (%): Liver: 97.3$\pm$1.3, Kidneys: 94.8$\pm$3.6, Spleen: 96.4$\pm$3.0, Pancreas: 80.9$\pm$10.1) at a rate of 1.6 seconds per image. Crucially, we are able to perform segmentation inference solely on CPU (no GPU required), thereby facilitating easy and widespread deployment of the model without specialist hardware.

arXiv:2207.09765공식 코드 · Papers with Code

ApHMM: Accelerating Profile Hidden Markov Models for Fast and Energy-Efficient Genome Analysis

Profile hidden Markov models (pHMMs) are widely employed in various bioinformatics applications to identify similarities between biological sequences, such as DNA or protein sequences. In pHMMs, sequences are represented as graph structures. These probabilities are subsequently used to compute the similarity score between a sequence and a pHMM graph. The Baum-Welch algorithm, a prevalent and highly accurate method, utilizes these probabilities to optimize and compute similarity scores. However, the Baum-Welch algorithm is computationally intensive, and existing solutions offer either software-only or hardware-only approaches with fixed pHMM designs. We identify an urgent need for a flexible, high-performance, and energy-efficient HW/SW co-design to address the major inefficiencies in the Baum-Welch algorithm for pHMMs. We introduce ApHMM, the first flexible acceleration framework designed to significantly reduce both computational and energy overheads associated with the Baum-Welch algorithm for pHMMs. ApHMM tackles the major inefficiencies in the Baum-Welch algorithm by 1) designing flexible hardware to accommodate various pHMM designs, 2) exploiting predictable data dependency patterns through on-chip memory with memoization techniques, 3) rapidly filtering out negligible computations using a hardware-based filter, and 4) minimizing redundant computations. ApHMM achieves substantial speedups of 15.55x - 260.03x, 1.83x - 5.34x, and 27.97x when compared to CPU, GPU, and FPGA implementations of the Baum-Welch algorithm, respectively. ApHMM outperforms state-of-the-art CPU implementations in three key bioinformatics applications: 1) error correction, 2) protein family search, and 3) multiple sequence alignment, by 1.29x - 59.94x, 1.03x - 1.75x, and 1.03x - 1.95x, respectively, while improving their energy efficiency by 64.24x - 115.46x, 1.75x, 1.96x.

arXiv:2207.08405

ORB-based SLAM accelerator on SoC FPGA

Simultaneous Localization and Mapping (SLAM) is one of the main components of autonomous navigation systems. With the increase in popularity of drones, autonomous navigation on low-power systems is seeing widespread application. Most SLAM algorithms are computationally intensive and struggle to run in real-time on embedded devices with reasonable accuracy. ORB-SLAM is an open-sourced feature-based SLAM that achieves high accuracy with reduced computational complexity. We propose an SoC based ORB-SLAM system that accelerates the computationally intensive visual feature extraction and matching on hardware. Our FPGA system based on a Zynq-family SoC runs 8.5x, 1.55x and 1.35x faster compared to an ARM CPU, Intel Desktop CPU, and a state-of-the-art FPGA system respectively, while averaging a 2x improvement in accuracy compared to prior work on FPGA.

arXiv:2207.07791

Associative Memory Based Experience Replay for Deep Reinforcement Learning

Experience replay is an essential component in deep reinforcement learning (DRL), which stores the experiences and generates experiences for the agent to learn in real time. Recently, prioritized experience replay (PER) has been proven to be powerful and widely deployed in DRL agents. However, implementing PER on traditional CPU or GPU architectures incurs significant latency overhead due to its frequent and irregular memory accesses. This paper proposes a hardware-software co-design approach to design an associative memory (AM) based PER, AMPER, with an AM-friendly priority sampling operation. AMPER replaces the widely-used time-costly tree-traversal-based priority sampling in PER while preserving the learning performance. Further, we design an in-memory computing hardware architecture based on AM to support AMPER by leveraging parallel in-memory search operations. AMPER shows comparable learning performance while achieving 55x to 270x latency improvement when running on the proposed hardware compared to the state-of-the-art PER running on GPU.

arXiv:2207.07736

Optimizing Data Collection in Deep Reinforcement Learning

Reinforcement learning (RL) workloads take a notoriously long time to train due to the large number of samples collected at run-time from simulators. Unfortunately, cluster scale-up approaches remain expensive, and commonly used CPU implementations of simulators induce high overhead when switching back and forth between GPU computations. We explore two optimizations that increase RL data collection efficiency by increasing GPU utilization: (1) GPU vectorization: parallelizing simulation on the GPU for increased hardware parallelism, and (2) simulator kernel fusion: fusing multiple simulation steps to run in a single GPU kernel launch to reduce global memory bandwidth requirements. We find that GPU vectorization can achieve up to $1024\times$ speedup over commonly used CPU simulators. We profile the performance of different implementations and show that for a simple simulator, ML compiler implementations (XLA) of GPU vectorization outperform a DNN framework (PyTorch) by $13.4\times$ by reducing CPU overhead from repeated Python to DL backend API calls. We show that simulator kernel fusion speedups with a simple simulator are $11.3\times$ and increase by up to $1024\times$ as simulator complexity increases in terms of memory bandwidth requirements. We show that the speedups from simulator kernel fusion are orthogonal and combinable with GPU vectorization, leading to a multiplicative speedup.

arXiv:2207.07886공식 코드 · Papers with Code

An Experimental Evaluation of Machine Learning Training on a Real Processing-in-Memory System

Training machine learning (ML) algorithms is a computationally intensive process, which is frequently memory-bound due to repeatedly accessing large training datasets. As a result, processor-centric systems (e.g., CPU, GPU) suffer from costly data movement between memory units and processing units, which consumes large amounts of energy and execution cycles. Memory-centric computing systems, i.e., with processing-in-memory (PIM) capabilities, can alleviate this data movement bottleneck. Our goal is to understand the potential of modern general-purpose PIM architectures to accelerate ML training. To do so, we (1) implement several representative classic ML algorithms (namely, linear regression, logistic regression, decision tree, K-Means clustering) on a real-world general-purpose PIM architecture, (2) rigorously evaluate and characterize them in terms of accuracy, performance and scaling, and (3) compare to their counterpart implementations on CPU and GPU. Our evaluation on a real memory-centric computing system with more than 2500 PIM cores shows that general-purpose PIM architectures can greatly accelerate memory-bound ML workloads, when the necessary operations and datatypes are natively supported by PIM hardware. For example, our PIM implementation of decision tree is $27\times$ faster than a state-of-the-art CPU version on an 8-core Intel Xeon, and $1.34\times$ faster than a state-of-the-art GPU version on an NVIDIA A100. Our K-Means clustering on PIM is $2.8\times$ and $3.2\times$ than state-of-the-art CPU and GPU versions, respectively. To our knowledge, our work is the first one to evaluate ML training on a real-world PIM architecture. We conclude with key observations, takeaways, and recommendations that can inspire users of ML workloads, programmers of PIM architectures, and hardware designers & architects of future memory-centric computing systems.

arXiv:2207.06920

Sub 8-Bit Quantization of Streaming Keyword Spotting Models for Embedded Chipsets

We propose a novel 2-stage sub 8-bit quantization aware training algorithm for all components of a 250K parameter feedforward, streaming, state-free keyword spotting model. For the 1st-stage, we adapt a recently proposed quantization technique using a non-linear transformation with tanh(.) on dense layer weights. In the 2nd-stage, we use linear quantization methods on the rest of the network, including other parameters (bias, gain, batchnorm), inputs, and activations. We conduct large scale experiments, training on 26,000 hours of de-identified production, far-field and near-field audio data (evaluating on 4,000 hours of data). We organize our results in two embedded chipset settings: a) with commodity ARM NEON instruction set and 8-bit containers, we present accuracy, CPU, and memory results using sub 8-bit weights (4, 5, 8-bit) and 8-bit quantization of rest of the network; b) with off-the-shelf neural network accelerators, for a range of weight bit widths (1 and 5-bit), while presenting accuracy results, we project reduction in memory utilization. In both configurations, our results show that the proposed algorithm can achieve: a) parity with a full floating point model's operating point on a detection error tradeoff (DET) curve in terms of false detection rate (FDR) at false rejection rate (FRR); b) significant reduction in compute and memory, yielding up to 3 times improvement in CPU consumption and more than 4 times improvement in memory consumption.