PyTorchKR Papers

최신 논문

총 5,575편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2003.10670

Real-time 3D object proposal generation and classification under limited processing resources

The task of detecting 3D objects is important to various robotic applications. The existing deep learning-based detection techniques have achieved impressive performance. However, these techniques are limited to run with a graphics processing unit (GPU) in a real-time environment. To achieve real-time 3D object detection with limited computational resources for robots, we propose an efficient detection method consisting of 3D proposal generation and classification. The proposal generation is mainly based on point segmentation, while the proposal classification is performed by a lightweight convolution neural network (CNN) model. To validate our method, KITTI datasets are utilized. The experimental results demonstrate the capability of proposed real-time 3D object detection method from the point cloud with a competitive performance of object recall and classification.

arXiv:2003.10555공식 코드 · Papers with Code

ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators

Masked language modeling (MLM) pre-training methods such as BERT corrupt the input by replacing some tokens with [MASK] and then train a model to reconstruct the original tokens. While they produce good results when transferred to downstream NLP tasks, they generally require large amounts of compute to be effective. As an alternative, we propose a more sample-efficient pre-training task called replaced token detection. Instead of masking the input, our approach corrupts it by replacing some tokens with plausible alternatives sampled from a small generator network. Then, instead of training a model that predicts the original identities of the corrupted tokens, we train a discriminative model that predicts whether each token in the corrupted input was replaced by a generator sample or not. Thorough experiments demonstrate this new pre-training task is more efficient than MLM because the task is defined over all input tokens rather than just the small subset that was masked out. As a result, the contextual representations learned by our approach substantially outperform the ones learned by BERT given the same model size, data, and compute. The gains are particularly strong for small models; for example, we train a model on one GPU for 4 days that outperforms GPT (trained using 30x more compute) on the GLUE natural language understanding benchmark. Our approach also works well at scale, where it performs comparably to RoBERTa and XLNet while using less than 1/4 of their compute and outperforms them when using the same amount of compute.

arXiv:2003.10520

Neural Game Engine: Accurate learning of generalizable forward models from pixels

Access to a fast and easily copied forward model of a game is essential for model-based reinforcement learning and for algorithms such as Monte Carlo tree search, and is also beneficial as a source of unlimited experience data for model-free algorithms. Learning forward models is an interesting and important challenge in order to address problems where a model is not available. Building upon previous work on the Neural GPU, this paper introduces the Neural Game Engine, as a way to learn models directly from pixels. The learned models are able to generalise to different size game levels to the ones they were trained on without loss of accuracy. Results on 10 deterministic General Video Game AI games demonstrate competitive performance, with many of the games models being learned perfectly both in terms of pixel predictions and reward predictions. The pre-trained models are available through the OpenAI Gym interface and are available publicly for future research here: \url{https://github.com/Bam4d/Neural-Game-Engine}

arXiv:2003.10959

Learning to Exploit Multiple Vision Modalities by Using Grafted Networks

Novel vision sensors such as thermal, hyperspectral, polarization, and event cameras provide information that is not available from conventional intensity cameras. An obstacle to using these sensors with current powerful deep neural networks is the lack of large labeled training datasets. This paper proposes a Network Grafting Algorithm (NGA), where a new front end network driven by unconventional visual inputs replaces the front end network of a pretrained deep network that processes intensity frames. The self-supervised training uses only synchronously-recorded intensity frames and novel sensor data to maximize feature similarity between the pretrained network and the grafted network. We show that the enhanced grafted network reaches competitive average precision (AP50) scores to the pretrained network on an object detection task using thermal and event camera datasets, with no increase in inference costs. Particularly, the grafted network driven by thermal frames showed a relative improvement of 49.11% over the use of intensity frames. The grafted front end has only 5--8% of the total parameters and can be trained in a few hours on a single GPU equivalent to 5% of the time that would be needed to train the entire object detector from labeled data. NGA allows new vision sensors to capitalize on previously pretrained powerful deep models, saving on training cost and widening a range of applications for novel sensors.

arXiv:2003.09696

PyCARL: A PyNN Interface for Hardware-Software Co-Simulation of Spiking Neural Network

We present PyCARL, a PyNN-based common Python programming interface for hardware-software co-simulation of spiking neural network (SNN). Through PyCARL, we make the following two key contributions. First, we provide an interface of PyNN to CARLsim, a computationally-efficient, GPU-accelerated and biophysically-detailed SNN simulator. PyCARL facilitates joint development of machine learning models and code sharing between CARLsim and PyNN users, promoting an integrated and larger neuromorphic community. Second, we integrate cycle-accurate models of state-of-the-art neuromorphic hardware such as TrueNorth, Loihi, and DynapSE in PyCARL, to accurately model hardware latencies that delay spikes between communicating neurons and degrade performance. PyCARL allows users to analyze and optimize the performance difference between software-only simulation and hardware-software co-simulation of their machine learning models. We show that system designers can also use PyCARL to perform design-space exploration early in the product development stage, facilitating faster time-to-deployment of neuromorphic products. We evaluate the memory usage and simulation time of PyCARL using functionality tests, synthetic SNNs, and realistic applications. Our results demonstrate that for large SNNs, PyCARL does not lead to any significant overhead compared to CARLsim. We also use PyCARL to analyze these SNNs for a state-of-the-art neuromorphic hardware and demonstrate a significant performance deviation from software-only simulations. PyCARL allows to evaluate and minimize such differences early during model development.

arXiv:2003.10142공식 코드 · Papers with Code

EPSNet: Efficient Panoptic Segmentation Network with Cross-layer Attention Fusion

Panoptic segmentation is a scene parsing task which unifies semantic segmentation and instance segmentation into one single task. However, the current state-of-the-art studies did not take too much concern on inference time. In this work, we propose an Efficient Panoptic Segmentation Network (EPSNet) to tackle the panoptic segmentation tasks with fast inference speed. Basically, EPSNet generates masks based on simple linear combination of prototype masks and mask coefficients. The light-weight network branches for instance segmentation and semantic segmentation only need to predict mask coefficients and produce masks with the shared prototypes predicted by prototype network branch. Furthermore, to enhance the quality of shared prototypes, we adopt a module called "cross-layer attention fusion module", which aggregates the multi-scale features with attention mechanism helping them capture the long-range dependencies between each other. To validate the proposed work, we have conducted various experiments on the challenging COCO panoptic dataset, which achieve highly promising performance with significantly faster inference speed (53ms on GPU).

arXiv:2003.10120공식 코드 · Papers with Code

Efficient Crowd Counting via Structured Knowledge Transfer

Crowd counting is an application-oriented task and its inference efficiency is crucial for real-world applications. However, most previous works relied on heavy backbone networks and required prohibitive run-time consumption, which would seriously restrict their deployment scopes and cause poor scalability. To liberate these crowd counting models, we propose a novel Structured Knowledge Transfer (SKT) framework, which fully exploits the structured knowledge of a well-trained teacher network to generate a lightweight but still highly effective student network. Specifically, it is integrated with two complementary transfer modules, including an Intra-Layer Pattern Transfer which sequentially distills the knowledge embedded in layer-wise features of the teacher network to guide feature learning of the student network and an Inter-Layer Relation Transfer which densely distills the cross-layer correlation knowledge of the teacher to regularize the student's feature evolutio Consequently, our student network can derive the layer-wise and cross-layer knowledge from the teacher network to learn compact yet effective features. Extensive evaluations on three benchmarks well demonstrate the effectiveness of our SKT for extensive crowd counting models. In particular, only using around $6\%$ of the parameters and computation cost of original models, our distilled VGG-based models obtain at least 6.5$\times$ speed-up on an Nvidia 1080 GPU and even achieve state-of-the-art performance. Our code and models are available at {\url{https://github.com/HCPLab-SYSU/SKT}}.

arXiv:1910.09020공식 코드 · Papers with Code

SneakySnake: A Fast and Accurate Universal Genome Pre-Alignment Filter for CPUs, GPUs, and FPGAs

Motivation: We introduce SneakySnake, a highly parallel and highly accurate pre-alignment filter that remarkably reduces the need for computationally costly sequence alignment. The key idea of SneakySnake is to reduce the approximate string matching (ASM) problem to the single net routing (SNR) problem in VLSI chip layout. In the SNR problem, we are interested in finding the optimal path that connects two terminals with the least routing cost on a special grid layout that contains obstacles. The SneakySnake algorithm quickly solves the SNR problem and uses the found optimal path to decide whether or not performing sequence alignment is necessary. Reducing the ASM problem into SNR also makes SneakySnake efficient to implement on CPUs, GPUs, and FPGAs. Results: SneakySnake significantly improves the accuracy of pre-alignment filtering by up to four orders of magnitude compared to the state-of-the-art pre-alignment filters, Shouji, GateKeeper, and SHD. For short sequences, SneakySnake accelerates Edlib (state-of-the-art implementation of Myers's bit-vector algorithm) and Parasail (state-of-the-art sequence aligner with a configurable scoring function), by up to 37.7x and 43.9x (>12x on average), respectively, with its CPU implementation, and by up to 413x and 689x (>400x on average), respectively, with FPGA and GPU acceleration. For long sequences, the CPU implementation of SneakySnake accelerates Parasail and KSW2 (sequence aligner of minimap2) by up to 979x (276.9x on average) and 91.7x (31.7x on average), respectively. As SneakySnake does not replace sequence alignment, users can still obtain all capabilities (e.g., configurable scoring functions) of the aligner of their choice, unlike existing acceleration efforts that sacrifice some aligner capabilities. Availability: https://github.com/CMU-SAFARI/SneakySnake

arXiv:2003.08436공식 코드 · Papers with Code

Collaborative Distillation for Ultra-Resolution Universal Style Transfer

Universal style transfer methods typically leverage rich representations from deep Convolutional Neural Network (CNN) models (e.g., VGG-19) pre-trained on large collections of images. Despite the effectiveness, its application is heavily constrained by the large model size to handle ultra-resolution images given limited memory. In this work, we present a new knowledge distillation method (named Collaborative Distillation) for encoder-decoder based neural style transfer to reduce the convolutional filters. The main idea is underpinned by a finding that the encoder-decoder pairs construct an exclusive collaborative relationship, which is regarded as a new kind of knowledge for style transfer models. Moreover, to overcome the feature size mismatch when applying collaborative distillation, a linear embedding loss is introduced to drive the student network to learn a linear embedding of the teacher's features. Extensive experiments show the effectiveness of our method when applied to different universal style transfer approaches (WCT and AdaIN), even if the model size is reduced by 15.5 times. Especially, on WCT with the compressed models, we achieve ultra-resolution (over 40 megapixels) universal style transfer on a 12GB GPU for the first time. Further experiments on optimization-based stylization scheme show the generality of our algorithm on different stylization paradigms. Our code and trained models are available at https://github.com/mingsun-tse/collaborative-distillation.

arXiv:2003.08394

Convergence of Artificial Intelligence and High Performance Computing on NSF-supported Cyberinfrastructure

Significant investments to upgrade and construct large-scale scientific facilities demand commensurate investments in R&D to design algorithms and computing approaches to enable scientific and engineering breakthroughs in the big data era. Innovative Artificial Intelligence (AI) applications have powered transformational solutions for big data challenges in industry and technology that now drive a multi-billion dollar industry, and which play an ever increasing role shaping human social patterns. As AI continues to evolve into a computing paradigm endowed with statistical and mathematical rigor, it has become apparent that single-GPU solutions for training, validation, and testing are no longer sufficient for computational grand challenges brought about by scientific facilities that produce data at a rate and volume that outstrip the computing capabilities of available cyberinfrastructure platforms. This realization has been driving the confluence of AI and high performance computing (HPC) to reduce time-to-insight, and to enable a systematic study of domain-inspired AI architectures and optimization schemes to enable data-driven discovery. In this article we present a summary of recent developments in this field, and describe specific advances that authors in this article are spearheading to accelerate and streamline the use of HPC platforms to design and apply accelerated AI algorithms in academia and industry.

arXiv:2003.08011

ContainerStress: Autonomous Cloud-Node Scoping Framework for Big-Data ML Use Cases

Deploying big-data Machine Learning (ML) services in a cloud environment presents a challenge to the cloud vendor with respect to the cloud container configuration sizing for any given customer use case. OracleLabs has developed an automated framework that uses nested-loop Monte Carlo simulation to autonomously scale any size customer ML use cases across the range of cloud CPU-GPU "Shapes" (configurations of CPUs and/or GPUs in Cloud containers available to end customers). Moreover, the OracleLabs and NVIDIA authors have collaborated on a ML benchmark study which analyzes the compute cost and GPU acceleration of any ML prognostic algorithm and assesses the reduction of compute cost in a cloud container comprising conventional CPUs and NVIDIA GPUs.

arXiv:2003.07543

KPNet: Towards Minimal Face Detector

The small receptive field and capacity of minimal neural networks limit their performance when using them to be the backbone of detectors. In this work, we find that the appearance feature of a generic face is discriminative enough for a tiny and shallow neural network to verify from the background. And the essential barriers behind us are 1) the vague definition of the face bounding box and 2) tricky design of anchor-boxes or receptive field. Unlike most top-down methods for joint face detection and alignment, the proposed KPNet detects small facial keypoints instead of the whole face by in a bottom-up manner. It first predicts the facial landmarks from a low-resolution image via the well-designed fine-grained scale approximation and scale adaptive soft-argmax operator. Finally, the precise face bounding boxes, no matter how we define it, can be inferred from the keypoints. Without any complex head architecture or meticulous network designing, the KPNet achieves state-of-the-art accuracy on generic face detection and alignment benchmarks with only $\sim1M$ parameters, which runs at 1000fps on GPU and is easy to perform real-time on most modern front-end chips.

arXiv:2003.06080

Coronary Artery Segmentation from Intravascular Optical Coherence Tomography Using Deep Capsules

The segmentation and analysis of coronary arteries from intravascular optical coherence tomography (IVOCT) is an important aspect of diagnosing and managing coronary artery disease. Current image processing methods are hindered by the time needed to generate expert-labelled datasets and the potential for bias during the analysis. Therefore, automated, robust, unbiased and timely geometry extraction from IVOCT, using image processing, would be beneficial to clinicians. With clinical application in mind, we aim to develop a model with a small memory footprint that is fast at inference time without sacrificing segmentation quality. Using a large IVOCT dataset of 12,011 expert-labelled images from 22 patients, we construct a new deep learning method based on capsules which automatically produces lumen segmentations. Our dataset contains images with both blood and light artefacts (22.8%), as well as metallic (23.1%) and bioresorbable stents (2.5%). We split the dataset into a training (70%), validation (20%) and test (10%) set and rigorously investigate design variations with respect to upsampling regimes and input selection. We show that our developments lead to a model, DeepCap, that is on par with state-of-the-art machine learning methods in terms of segmentation quality and robustness, while using as little as 12% of the parameters. This enables DeepCap to have per image inference times up to 70% faster on GPU and up to 95% faster on CPU compared to other state-of-the-art models. DeepCap is a robust automated segmentation tool that can aid clinicians to extract unbiased geometrical data from IVOCT.

arXiv:2003.05622

Distributed Hierarchical GPU Parameter Server for Massive Scale Deep Learning Ads Systems

Neural networks of ads systems usually take input from multiple resources, e.g., query-ad relevance, ad features and user portraits. These inputs are encoded into one-hot or multi-hot binary features, with typically only a tiny fraction of nonzero feature values per example. Deep learning models in online advertising industries can have terabyte-scale parameters that do not fit in the GPU memory nor the CPU main memory on a computing node. For example, a sponsored online advertising system can contain more than $10^{11}$ sparse features, making the neural network a massive model with around 10 TB parameters. In this paper, we introduce a distributed GPU hierarchical parameter server for massive scale deep learning ads systems. We propose a hierarchical workflow that utilizes GPU High-Bandwidth Memory, CPU main memory and SSD as 3-layer hierarchical storage. All the neural network training computations are contained in GPUs. Extensive experiments on real-world data confirm the effectiveness and the scalability of the proposed system. A 4-node hierarchical GPU parameter server can train a model more than 2X faster than a 150-node in-memory distributed parameter server in an MPI cluster. In addition, the price-performance ratio of our proposed system is 4-9 times better than an MPI-cluster solution.

arXiv:2003.05926공식 코드 · Papers with Code

Learning distributed representations of graphs with Geo2DR

We present Geo2DR (Geometric to Distributed Representations), a GPU ready Python library for unsupervised learning on graph-structured data using discrete substructure patterns and neural language models. It contains efficient implementations of popular graph decomposition algorithms and neural language models in PyTorch which can be combined to learn representations of graphs using the distributive hypothesis. Furthermore, Geo2DR comes with general data processing and loading methods to bring substantial speed-up in the training of the neural language models. Through this we provide a modular set of tools and methods to quickly construct systems capable of learning distributed representations of graphs. This is useful for replication of existing methods, modification, or development of completely new methods. This paper serves to present the Geo2DR library and perform a comprehensive comparative analysis of existing methods re-implemented using Geo2DR across widely used graph classification benchmarks. Geo2DR displays a high reproducibility of results in published methods and interoperability with other libraries useful for distributive language modelling.

arXiv:2002.05200공식 코드 · Papers with Code

LOGAN: High-Performance GPU-Based X-Drop Long-Read Alignment

Pairwise sequence alignment is one of the most computationally intensive kernels in genomic data analysis, accounting for more than 90% of the runtime for key bioinformatics applications. This method is particularly expensive for third-generation sequences due to the high computational cost of analyzing sequences of length between 1Kb and 1Mb. Given the quadratic overhead of exact pairwise algorithms for long alignments, the community primarily relies on approximate algorithms that search only for high-quality alignments and stop early when one is not found. In this work, we present the first GPU optimization of the popular X-drop alignment algorithm, that we named LOGAN. Results show that our high-performance multi-GPU implementation achieves up to 181.6 GCUPS and speed-ups up to 6.6x and 30.7x using 1 and 6 NVIDIA Tesla V100, respectively, over the state-of-the-art software running on two IBM Power9 processors using 168 CPU threads, with equivalent accuracy. We also demonstrate a 2.3x LOGAN speed-up versus ksw2, a state-of-art vectorized algorithm for sequence alignment implemented in minimap2, a long-read mapping software. To highlight the impact of our work on a real-world application, we couple LOGAN with a many-to-many long-read alignment software called BELLA, and demonstrate that our implementation improves the overall BELLA runtime by up to 10.6x. Finally, we adapt the Roofline model for LOGAN and demonstrate that our implementation is near-optimal on the NVIDIA Tesla V100s.

arXiv:2003.04617공식 코드 · Papers with Code

Differentiate Everything with a Reversible Embeded Domain-Specific Language

Reverse-mode automatic differentiation (AD) suffers from the issue of having too much space overhead to trace back intermediate computational states for back-propagation. The traditional method to trace back states is called checkpointing that stores intermediate states into a global stack and restore state through either stack pop or re-computing. The overhead of stack manipulations and re-computing makes the general purposed (not tensor-based) AD engines unable to meet many industrial needs. Instead of checkpointing, we propose to use reverse computing to trace back states by designing and implementing a reversible programming eDSL, where a program can be executed bi-directionally without implicit stack operations. The absence of implicit stack operations makes the program compatible with existing compiler features, including utilizing existing optimization passes and compiling the code as GPU kernels. We implement AD for sparse matrix operations and some machine learning applications to show that our framework has the state-of-the-art performance.

arXiv:2003.04294공식 코드 · Papers with Code

Optimizing Streaming Parallelism on Heterogeneous Many-Core Architectures: A Machine Learning Based Approach

This article presents an automatic approach to quickly derive a good solution for hardware resource partition and task granularity for task-based parallel applications on heterogeneous many-core architectures. Our approach employs a performance model to estimate the resulting performance of the target application under a given resource partition and task granularity configuration. The model is used as a utility to quickly search for a good configuration at runtime. Instead of hand-crafting an analytical model that requires expert insights into low-level hardware details, we employ machine learning techniques to automatically learn it. We achieve this by first learning a predictive model offline using training programs. The learnt model can then be used to predict the performance of any unseen program at runtime. We apply our approach to 39 representative parallel applications and evaluate it on two representative heterogeneous many-core platforms: a CPU-XeonPhi platform and a CPU-GPU platform. Compared to the single-stream version, our approach achieves, on average, a 1.6x and 1.1x speedup on the XeonPhi and the GPU platform, respectively. These results translate to over 93% of the performance delivered by a theoretically perfect predictor.

arXiv:2003.03913

FarSee-Net: Real-Time Semantic Segmentation by Efficient Multi-scale Context Aggregation and Feature Space Super-resolution

Real-time semantic segmentation is desirable in many robotic applications with limited computation resources. One challenge of semantic segmentation is to deal with the object scale variations and leverage the context. How to perform multi-scale context aggregation within limited computation budget is important. In this paper, firstly, we introduce a novel and efficient module called Cascaded Factorized Atrous Spatial Pyramid Pooling (CF-ASPP). It is a lightweight cascaded structure for Convolutional Neural Networks (CNNs) to efficiently leverage context information. On the other hand, for runtime efficiency, state-of-the-art methods will quickly decrease the spatial size of the inputs or feature maps in the early network stages. The final high-resolution result is usually obtained by non-parametric up-sampling operation (e.g. bilinear interpolation). Differently, we rethink this pipeline and treat it as a super-resolution process. We use optimized super-resolution operation in the up-sampling step and improve the accuracy, especially in sub-sampled input image scenario for real-time applications. By fusing the above two improvements, our methods provide better latency-accuracy trade-off than the other state-of-the-art methods. In particular, we achieve 68.4% mIoU at 84 fps on the Cityscapes test set with a single Nivida Titan X (Maxwell) GPU card. The proposed module can be plugged into any feature extraction CNN and benefits from the CNN structure development.

arXiv:2003.03771공식 코드 · Papers with Code

Pixel-in-Pixel Net: Towards Efficient Facial Landmark Detection in the Wild

Recently, heatmap regression models have become popular due to their superior performance in locating facial landmarks. However, three major problems still exist among these models: (1) they are computationally expensive; (2) they usually lack explicit constraints on global shapes; (3) domain gaps are commonly present. To address these problems, we propose Pixel-in-Pixel Net (PIPNet) for facial landmark detection. The proposed model is equipped with a novel detection head based on heatmap regression, which conducts score and offset predictions simultaneously on low-resolution feature maps. By doing so, repeated upsampling layers are no longer necessary, enabling the inference time to be largely reduced without sacrificing model accuracy. Besides, a simple but effective neighbor regression module is proposed to enforce local constraints by fusing predictions from neighboring landmarks, which enhances the robustness of the new detection head. To further improve the cross-domain generalization capability of PIPNet, we propose self-training with curriculum. This training strategy is able to mine more reliable pseudo-labels from unlabeled data across domains by starting with an easier task, then gradually increasing the difficulty to provide more precise labels. Extensive experiments demonstrate the superiority of PIPNet, which obtains state-of-the-art results on three out of six popular benchmarks under the supervised setting. The results on two cross-domain test sets are also consistently improved compared to the baselines. Notably, our lightweight version of PIPNet runs at 35.7 FPS and 200 FPS on CPU and GPU, respectively, while still maintaining a competitive accuracy to state-of-the-art methods. The code of PIPNet is available at https://github.com/jhb86253817/PIPNet.