PyTorchKR Papers

최신 논문

총 5,575편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2010.02871

LETI: Latency Estimation Tool and Investigation of Neural Networks inference on Mobile GPU

A lot of deep learning applications are desired to be run on mobile devices. Both accuracy and inference time are meaningful for a lot of them. While the number of FLOPs is usually used as a proxy for neural network latency, it may be not the best choice. In order to obtain a better approximation of latency, research community uses look-up tables of all possible layers for latency calculation for the final prediction of the inference on mobile CPU. It requires only a small number of experiments. Unfortunately, on mobile GPU this method is not applicable in a straight-forward way and shows low precision. In this work, we consider latency approximation on mobile GPU as a data and hardware-specific problem. Our main goal is to construct a convenient latency estimation tool for investigation(LETI) of neural network inference and building robust and accurate latency prediction models for each specific task. To achieve this goal, we build open-source tools which provide a convenient way to conduct massive experiments on different target devices focusing on mobile GPU. After evaluation of the dataset, we learn the regression model on experimental data and use it for future latency prediction and analysis. We experimentally demonstrate the applicability of such an approach on a subset of popular NAS-Benchmark 101 dataset and also evaluate the most popular neural network architectures for two mobile GPUs. As a result, we construct latency prediction model with good precision on the target evaluation subset. We consider LETI as a useful tool for neural architecture search or massive latency evaluation. The project is available at https://github.com/leti-ai

arXiv:2010.02416

Efficient Inference For Neural Machine Translation

Large Transformer models have achieved state-of-the-art results in neural machine translation and have become standard in the field. In this work, we look for the optimal combination of known techniques to optimize inference speed without sacrificing translation quality. We conduct an empirical study that stacks various approaches and demonstrates that combination of replacing decoder self-attention with simplified recurrent units, adopting a deep encoder and a shallow decoder architecture and multi-head attention pruning can achieve up to 109% and 84% speedup on CPU and GPU respectively and reduce the number of parameters by 25% while maintaining the same translation quality in terms of BLEU.

arXiv:2010.01526

NLP Service APIs and Models for Efficient Registration of New Clients

State-of-the-art NLP inference uses enormous neural architectures and models trained for GPU-months, well beyond the reach of most consumers of NLP. This has led to one-size-fits-all public API-based NLP service models by major AI companies, serving large numbers of clients. Neither (hardware deficient) clients nor (heavily subscribed) servers can afford traditional fine tuning. Many clients own little or no labeled data. We initiate a study of adaptation of centralized NLP services to clients, and present one practical and lightweight approach. Each client uses an unsupervised, corpus-based sketch to register to the service. The server uses an auxiliary network to map the sketch to an abstract vector representation, which then informs the main labeling network. When a new client registers with its sketch, it gets immediate accuracy benefits. We demonstrate the success of the proposed architecture using sentiment labeling, NER, and predictive language modeling

arXiv:2010.01668

Diagonal Memory Optimisation for Machine Learning on Micro-controllers

As machine learning spreads into more and more application areas, micro controllers and low power CPUs are increasingly being used to perform inference with machine learning models. The capability to deploy onto these limited hardware targets is enabling machine learning models to be used across a diverse range of new domains. Optimising the inference process on these targets poses different challenges from either desktop CPU or GPU implementations, where the small amounts of RAM available on these targets sets limits on size of models which can be executed. Analysis of the memory use patterns of eleven machine learning models was performed. Memory load and store patterns were observed using a modified version of the Valgrind debugging tool, identifying memory areas holding values necessary for the calculation as inference progressed. These analyses identified opportunities optimise the memory use of these models by overlapping the input and output buffers of individual tensor operations. Three methods are presented which can calculate the safe overlap of input and output buffers for tensor operations. Ranging from a computationally expensive approach with the ability to operate on compiled layer operations, to a versatile analytical solution which requires access to the original source code of the layer. The diagonal memory optimisation technique is described and shown to achieve memory savings of up to 34.5% when applied to eleven common models. Micro-controller targets are identified where it is only possible to deploy some models if diagonal memory optimisation is used.

arXiv:2010.01652공식 코드 · Papers with Code

FORK: A Forward-Looking Actor For Model-Free Reinforcement Learning

In this paper, we propose a new type of Actor, named forward-looking Actor or FORK for short, for Actor-Critic algorithms. FORK can be easily integrated into a model-free Actor-Critic algorithm. Our experiments on six Box2D and MuJoCo environments with continuous state and action spaces demonstrate significant performance improvement FORK can bring to the state-of-the-art algorithms. A variation of FORK can further solve Bipedal-WalkerHardcore in as few as four hours using a single GPU.

arXiv:2010.02193공식 코드 · Papers with Code

Mastering Atari with Discrete World Models

Intelligent agents need to generalize from past experience to achieve goals in complex environments. World models facilitate such generalization and allow learning behaviors from imagined outcomes to increase sample-efficiency. While learning world models from image inputs has recently become feasible for some tasks, modeling Atari games accurately enough to derive successful behaviors has remained an open challenge for many years. We introduce DreamerV2, a reinforcement learning agent that learns behaviors purely from predictions in the compact latent space of a powerful world model. The world model uses discrete representations and is trained separately from the policy. DreamerV2 constitutes the first agent that achieves human-level performance on the Atari benchmark of 55 tasks by learning behaviors inside a separately trained world model. With the same computational budget and wall-clock time, Dreamer V2 reaches 200M frames and surpasses the final performance of the top single-GPU agents IQN and Rainbow. DreamerV2 is also applicable to tasks with continuous actions, where it learns an accurate world model of a complex humanoid robot and solves stand-up and walking from only pixel inputs.

arXiv:2010.02164공식 코드 · Papers with Code

A Streaming Approach For Efficient Batched Beam Search

We propose an efficient batching strategy for variable-length decoding on GPU architectures. During decoding, when candidates terminate or are pruned according to heuristics, our streaming approach periodically "refills" the batch before proceeding with a selected subset of candidates. We apply our method to variable-width beam search on a state-of-the-art machine translation model. Our method decreases runtime by up to 71% compared to a fixed-width beam search baseline and 17% compared to a variable-width baseline, while matching baselines' BLEU. Finally, experiments show that our method can speed up decoding in other domains, such as semantic and syntactic parsing.

arXiv:2010.01895

Joint Scene and Object Tracking for Cost-Effective Augmented Reality Assisted Patient Positioning in Radiation Therapy

BACKGROUND AND OBJECTIVE: The research done in the field of Augmented Reality (AR) for patient positioning in radiation therapy is scarce. We propose an efficient and cost-effective algorithm for tracking the scene and the patient to interactively assist the patient's positioning process by providing visual feedback to the operator. Up to our knowledge, this is the first framework that can be employed for mobile interactive AR to guide patient positioning. METHODS: We propose a point cloud processing method that combined with a fiducial marker-mapper algorithm and the generalized ICP algorithm tracks the patient and the camera precisely and efficiently only using the CPU unit. The alignment between the 3D reference model and body marker map is calculated employing an efficient body reconstruction algorithm. RESULTS: Our quantitative evaluation shows that the proposed method achieves a translational and rotational error of 4.17 mm/0.82 deg at 9 fps. Furthermore, the qualitative results demonstrate the usefulness of our algorithm in patient positioning on different human subjects. CONCLUSION: Since our algorithm achieves a relatively high frame rate and accuracy employing a regular laptop (without the usage of a dedicated GPU), it is a very cost-effective AR-based patient positioning method. It also opens the way for other researchers by introducing a framework that could be improved upon for better mobile interactive AR patient positioning solutions in the future.

arXiv:2003.02978공식 코드 · Papers with Code

Fast and Accurate Retrieval of Methane Concentration from Imaging Spectrometer Data Using Sparsity Prior

The strong radiative forcing by atmospheric methane has stimulated interest in identifying natural and anthropogenic sources of this potent greenhouse gas. Point sources are important targets for quantification, and anthropogenic targets have potential for emissions reduction. Methane point source plume detection and concentration retrieval have been previously demonstrated using data from the Airborne Visible InfraRed Imaging Spectrometer Next Generation (AVIRIS-NG). Current quantitative methods have tradeoffs between computational requirements and retrieval accuracy, creating obstacles for processing real-time data or large datasets from flight campaigns. We present a new computationally efficient algorithm that applies sparsity and an albedo correction to matched filter retrieval of trace gas concentration-pathlength. The new algorithm was tested using AVIRIS-NG data acquired over several point source plumes in Ahmedabad, India. The algorithm was validated using simulated AVIRIS-NG data including synthetic plumes of known methane concentration. Sparsity and albedo correction together reduced the root mean squared error of retrieved methane concentration-pathlength enhancement by 60.7% compared with a previous robust matched filter method. Background noise was reduced by a factor of 2.64. The new algorithm was able to process the entire 300 flightline 2016 AVIRIS-NG India campaign in just over 8 hours on a desktop computer with GPU acceleration.

arXiv:2009.14788공식 코드 · Papers with Code

TorchRadon: Fast Differentiable Routines for Computed Tomography

This work presents TorchRadon -- an open source CUDA library which contains a set of differentiable routines for solving computed tomography (CT) reconstruction problems. The library is designed to help researchers working on CT problems to combine deep learning and model-based approaches. The package is developed as a PyTorch extension and can be seamlessly integrated into existing deep learning training code. Compared to the existing Astra Toolbox, TorchRadon is up to 125 faster. The operators implemented by TorchRadon allow the computation of gradients using PyTorch backward(), and can therefore be easily inserted inside existing neural networks architectures. Because of its speed and GPU support, TorchRadon can also be effectively used as a fast backend for the implementation of iterative algorithms. This paper presents the main functionalities of the library, compares results with existing libraries and provides examples of usage.

arXiv:2009.14684

Benchmark for Anonymous Video Analytics

Out-of-home audience measurement aims to count and characterize the people exposed to advertising content in the physical world. While audience measurement solutions based on computer vision are of increasing interest, no commonly accepted benchmark exists to evaluate and compare their performance. In this paper, we propose the first benchmark for digital out-of-home audience measurement that evaluates the vision-based tasks of audience localization and counting, and audience demographics. The benchmark is composed of a novel, dataset captured at multiple locations and a set of performance measures. Using the benchmark, we present an in-depth comparison of eight open-source algorithms on four hardware platforms with GPU and CPU-optimized inferences and of two commercial off-the-shelf solutions for localization, count, age, and gender estimation. This benchmark and related open-source codes are available at http://ava.eecs.qmul.ac.uk.

arXiv:2009.14554

One Reflection Suffice

Orthogonal weight matrices are used in many areas of deep learning. Much previous work attempt to alleviate the additional computational resources it requires to constrain weight matrices to be orthogonal. One popular approach utilizes *many* Householder reflections. The only practical drawback is that many reflections cause low GPU utilization. We mitigate this final drawback by proving that *one* reflection is sufficient, if the reflection is computed by an auxiliary neural network.

arXiv:2009.14468공식 코드 · Papers with Code

Learning Image-adaptive 3D Lookup Tables for High Performance Photo Enhancement in Real-time

Recent years have witnessed the increasing popularity of learning based methods to enhance the color and tone of photos. However, many existing photo enhancement methods either deliver unsatisfactory results or consume too much computational and memory resources, hindering their application to high-resolution images (usually with more than 12 megapixels) in practice. In this paper, we learn image-adaptive 3-dimensional lookup tables (3D LUTs) to achieve fast and robust photo enhancement. 3D LUTs are widely used for manipulating color and tone of photos, but they are usually manually tuned and fixed in camera imaging pipeline or photo editing tools. We, for the first time to our best knowledge, propose to learn 3D LUTs from annotated data using pairwise or unpaired learning. More importantly, our learned 3D LUT is image-adaptive for flexible photo enhancement. We learn multiple basis 3D LUTs and a small convolutional neural network (CNN) simultaneously in an end-to-end manner. The small CNN works on the down-sampled version of the input image to predict content-dependent weights to fuse the multiple basis 3D LUTs into an image-adaptive one, which is employed to transform the color and tone of source images efficiently. Our model contains less than 600K parameters and takes less than 2 ms to process an image of 4K resolution using one Titan RTX GPU. While being highly efficient, our model also outperforms the state-of-the-art photo enhancement methods by a large margin in terms of PSNR, SSIM and a color difference metric on two publically available benchmark datasets.

arXiv:2009.13716

Grow-Push-Prune: aligning deep discriminants for effective structural network compression

Most of today's popular deep architectures are hand-engineered to be generalists. However, this design procedure usually leads to massive redundant, useless, or even harmful features for specific tasks. Unnecessarily high complexities render deep nets impractical for many real-world applications, especially those without powerful GPU support. In this paper, we attempt to derive task-dependent compact models from a deep discriminant analysis perspective. We propose an iterative and proactive approach for classification tasks which alternates between (1) a pushing step, with an objective to simultaneously maximize class separation, penalize co-variances, and push deep discriminants into alignment with a compact set of neurons, and (2) a pruning step, which discards less useful or even interfering neurons. Deconvolution is adopted to reverse 'unimportant' filters' effects and recover useful contributing sources. A simple network growing strategy based on the basic Inception module is proposed for challenging tasks requiring larger capacity than what the base net can offer. Experiments on the MNIST, CIFAR10, and ImageNet datasets demonstrate our approach's efficacy. On ImageNet, by pushing and pruning our grown Inception-88 model, we achieve more accurate models than Inception nets generated during growing, residual nets, and popular compact nets at similar sizes. We also show that our grown Inception nets (without hard-coded dimension alignment) clearly outperform residual nets of similar complexities.

arXiv:2009.13062

Accelerating Multi-Model Inference by Merging DNNs of Different Weights

Standardized DNN models that have been proved to perform well on machine learning tasks are widely used and often adopted as-is to solve downstream tasks, forming the transfer learning paradigm. However, when serving multiple instances of such DNN models from a cluster of GPU servers, existing techniques to improve GPU utilization such as batching are inapplicable because models often do not share weights due to fine-tuning. We propose NetFuse, a technique of merging multiple DNN models that share the same architecture but have different weights and different inputs. NetFuse is made possible by replacing operations with more general counterparts that allow a set of weights to be associated with only a certain set of inputs. Experiments on ResNet-50, ResNeXt-50, BERT, and XLNet show that NetFuse can speed up DNN inference time up to 3.6x on a NVIDIA V100 GPU, and up to 3.0x on a TITAN Xp GPU when merging 32 model instances, while only using up a small additional amount of GPU memory.

arXiv:2009.10924

FusionStitching: Boosting Memory Intensive Computations for Deep Learning Workloads

We show in this work that memory intensive computations can result in severe performance problems due to off-chip memory access and CPU-GPU context switch overheads in a wide range of deep learning models. For this problem, current just-in-time (JIT) kernel fusion and code generation techniques have limitations, such as rough fusion plan exploration strategies and limited code generation ability. We propose FusionStitching, a deep learning compiler capable of fusing memory intensive operators, with varied data dependencies and non-homogeneous parallelism, into large GPU kernels to reduce global memory access and context switch overhead automatically. FusionStitching widens the range of operation combinations that fusion can target beyond previous JIT works by introducing data reuse of intermediate values. It explores large fusion spaces to decide optimal fusion plans with considerations of memory access costs, kernel calls and resource usage constraints. FusionStitching tunes the optimal stitching scheme with a domain-specific cost model efficiently. Experimental results show that FusionStitching can reach up to 2.21x speedup compared to state-of-the-art, with 1.45x on average. Besides these experimental results, we integrated our approach into a compiler product and deployed it onto a production cluster for AI workloads with thousands of GPUs. The system has been in operation for more than 4 months and saves 7,000 GPU hours on average for approximately 30,000 tasks per month.

arXiv:2009.11112공식 코드 · Papers with Code

ANNdotNET -- deep learning tool on .NET Platform

ANNdotNET is an open source project for deep learning written in C# with ability to create, train, evaluate and export deep learning models. The project consists of the Graphical User Interface module capable to visually prepare data, fine tune hyper-parameters, design network architecture, evaluate and test trained models. The ANNdotNET introduces the Visual Network Designer, (VND) for visually design almost any sequential deep learning network. Beside VND, ANNdotNET implements Machine Learning Engine, (MLE) based on CNTK - deep learning framework, with ability to train and evaluate models on GPU. For model evaluation ANNdotNET contains rich set of visual and descriptive performance parameters, history of the training process and set of export/deployment options. The advantage of using ANNdotNET over the classic code based ML approach is more focus on deep learning network design and training process instead of focusing on coding and debugging. It is ideal for engineers not familiar with supported programming languages. The project is hosted at github.com/bhrnjica/anndotnet.

arXiv:2009.10370

Visual Methods for Sign Language Recognition: A Modality-Based Review

Sign language visual recognition from continuous multi-modal streams is still one of the most challenging fields. Recent advances in human actions recognition are exploiting the ascension of GPU-based learning from massive data, and are getting closer to human-like performances. They are then prone to creating interactive services for the deaf and hearing-impaired communities. A population that is expected to grow considerably in the years to come. This paper aims at reviewing the human actions recognition literature with the sign-language visual understanding as a scope. The methods analyzed will be mainly organized according to the different types of unimodal inputs exploited, their relative multi-modal combinations and pipeline steps. In each section, we will detail and compare the related datasets, approaches then distinguish the still open contribution paths suitable for the creation of sign language related services. Special attention will be paid to the approaches and commercial solutions handling facial expressions and continuous signing.

arXiv:2009.09209

MSR-DARTS: Minimum Stable Rank of Differentiable Architecture Search

In neural architecture search (NAS), differentiable architecture search (DARTS) has recently attracted much attention due to its high efficiency. It defines an over-parameterized network with mixed edges, each of which represents all operator candidates, and jointly optimizes the weights of the network and its architecture in an alternating manner. However, this method finds a model with the weights converging faster than the others, and such a model with fastest convergence often leads to overfitting. Accordingly, the resulting model cannot always be well-generalized. To overcome this problem, we propose a method called minimum stable rank DARTS (MSR-DARTS), for finding a model with the best generalization error by replacing architecture optimization with the selection process using the minimum stable rank criterion. Specifically, a convolution operator is represented by a matrix, and MSR-DARTS selects the one with the smallest stable rank. We evaluated MSR-DARTS on CIFAR-10 and ImageNet datasets. It achieves an error rate of 2.54% with 4.0M parameters within 0.3 GPU-days on CIFAR-10, and a top-1 error rate of 23.9% on ImageNet. The official code is available at https://github.com/mtaecchhi/msrdarts.git.

arXiv:2009.08924

Multi-Resolution Graph Neural Network for Large-Scale Pointcloud Segmentation

In this paper, we propose a multi-resolution deep-learning architecture to semantically segment dense large-scale pointclouds. Dense pointcloud data require a computationally expensive feature encoding process before semantic segmentation. Previous work has used different approaches to drastically downsample from the original pointcloud so common computing hardware can be utilized. While these approaches can relieve the computation burden to some extent, they are still limited in their processing capability for multiple scans. We present MuGNet, a memory-efficient, end-to-end graph neural network framework to perform semantic segmentation on large-scale pointclouds. We reduce the computation demand by utilizing a graph neural network on the preformed pointcloud graphs and retain the precision of the segmentation with a bidirectional network that fuses feature embedding at different resolutions. Our framework has been validated on benchmark datasets including Stanford Large-Scale 3D Indoor Spaces Dataset(S3DIS) and Virtual KITTI Dataset. We demonstrate that our framework can process up to 45 room scans at once on a single 11 GB GPU while still surpassing other graph-based solutions for segmentation on S3DIS with an 88.5\% (+3\%) overall accuracy and 69.8\% (+7.7\%) mIOU accuracy.