PyTorchKR Papers

최신 논문

총 5,575편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1611.05319

Guidefill: GPU Accelerated, Artist Guided Geometric Inpainting for 3D Conversion

The conversion of traditional film into stereo 3D has become an important problem in the past decade. One of the main bottlenecks is a disocclusion step, which in commercial 3D conversion is usually done by teams of artists armed with a toolbox of inpainting algorithms. A current difficulty in this is that most available algorithms are either too slow for interactive use, or provide no intuitive means for users to tweak the output. In this paper we present a new fast inpainting algorithm based on transporting along automatically detected splines, which the user may edit. Our algorithm is implemented on the GPU and fills the inpainting domain in successive shells that adapt their shape on the fly. In order to allocate GPU resources as efficiently as possible, we propose a parallel algorithm to track the inpainting interface as it evolves, ensuring that no resources are wasted on pixels that are not currently being worked on. Theoretical analysis of the time and processor complexiy of our algorithm without and with tracking (as well as numerous numerical experiments) demonstrate the merits of the latter. Our transport mechanism is similar to the one used in coherence transport, but improves upon it by corrected a "kinking" phenomena whereby extrapolated isophotes may bend at the boundary of the inpainting domain. Theoretical results explaining this phenomena and its resolution are presented. Although our method ignores texture, in many cases this is not a problem due to the thin inpainting domains in 3D conversion. Experimental results show that our method can achieve a visual quality that is competitive with the state-of-the-art while maintaining interactive speeds and providing the user with an intuitive interface to tweak the results.

arXiv:1708.00187

Real-time Deep Video Deinterlacing

Interlacing is a widely used technique, for television broadcast and video recording, to double the perceived frame rate without increasing the bandwidth. But it presents annoying visual artifacts, such as flickering and silhouette "serration," during the playback. Existing state-of-the-art deinterlacing methods either ignore the temporal information to provide real-time performance but lower visual quality, or estimate the motion for better deinterlacing but with a trade-off of higher computational cost. In this paper, we present the first and novel deep convolutional neural networks (DCNNs) based method to deinterlace with high visual quality and real-time performance. Unlike existing models for super-resolution problems which relies on the translation-invariant assumption, our proposed DCNN model utilizes the temporal information from both the odd and even half frames to reconstruct only the missing scanlines, and retains the given odd and even scanlines for producing the full deinterlaced frames. By further introducing a layer-sharable architecture, our system can achieve real-time performance on a single GPU. Experiments shows that our method outperforms all existing methods, in terms of reconstruction accuracy and computational performance.

arXiv:1708.00079

Towards the Success Rate of One: Real-time Unconstrained Salient Object Detection

In this work, we propose an efficient and effective approach for unconstrained salient object detection in images using deep convolutional neural networks. Instead of generating thousands of candidate bounding boxes and refining them, our network directly learns to generate the saliency map containing the exact number of salient objects. During training, we convert the ground-truth rectangular boxes to Gaussian distributions that better capture the ROI regarding individual salient objects. During inference, the network predicts Gaussian distributions centered at salient objects with an appropriate covariance, from which bounding boxes are easily inferred. Notably, our network performs saliency map prediction without pixel-level annotations, salient object detection without object proposals, and salient object subitizing simultaneously, all in a single pass within a unified framework. Extensive experiments show that our approach outperforms existing methods on various datasets by a large margin, and achieves more than 100 fps with VGG16 network on a single GPU during inference.

arXiv:1708.00783

InfiniTAM v3: A Framework for Large-Scale 3D Reconstruction with Loop Closure

Volumetric models have become a popular representation for 3D scenes in recent years. One breakthrough leading to their popularity was KinectFusion, which focuses on 3D reconstruction using RGB-D sensors. However, monocular SLAM has since also been tackled with very similar approaches. Representing the reconstruction volumetrically as a TSDF leads to most of the simplicity and efficiency that can be achieved with GPU implementations of these systems. However, this representation is memory-intensive and limits applicability to small-scale reconstructions. Several avenues have been explored to overcome this. With the aim of summarizing them and providing for a fast, flexible 3D reconstruction pipeline, we propose a new, unifying framework called InfiniTAM. The idea is that steps like camera tracking, scene representation and integration of new data can easily be replaced and adapted to the user's needs. This report describes the technical implementation details of InfiniTAM v3, the third version of our InfiniTAM system. We have added various new features, as well as making numerous enhancements to the low-level code that significantly improve our camera tracking performance. The new features that we expect to be of most interest are (i) a robust camera tracking module; (ii) an implementation of Glocker et al.'s keyframe-based random ferns camera relocaliser; (iii) a novel approach to globally-consistent TSDF-based reconstruction, based on dividing the scene into rigid submaps and optimising the relative poses between them; and (iv) an implementation of Keller et al.'s surfel-based reconstruction approach.

arXiv:1703.07814공식 코드 · Papers with Code

R-C3D: Region Convolutional 3D Network for Temporal Activity Detection

We address the problem of activity detection in continuous, untrimmed video streams. This is a difficult task that requires extracting meaningful spatio-temporal features to capture activities, accurately localizing the start and end times of each activity. We introduce a new model, Region Convolutional 3D Network (R-C3D), which encodes the video streams using a three-dimensional fully convolutional network, then generates candidate temporal regions containing activities, and finally classifies selected regions into specific activities. Computation is saved due to the sharing of convolutional features between the proposal and the classification pipelines. The entire model is trained end-to-end with jointly optimized localization and classification losses. R-C3D is faster than existing methods (569 frames per second on a single Titan X Maxwell GPU) and achieves state-of-the-art results on THUMOS'14. We further demonstrate that our model is a general activity detection framework that does not rely on assumptions about particular dataset properties by evaluating our approach on ActivityNet and Charades. Our code is available at http://ai.bu.edu/r-c3d/.

arXiv:1703.06189공식 코드 · Papers with Code

TURN TAP: Temporal Unit Regression Network for Temporal Action Proposals

Temporal Action Proposal (TAP) generation is an important problem, as fast and accurate extraction of semantically important (e.g. human actions) segments from untrimmed videos is an important step for large-scale video analysis. We propose a novel Temporal Unit Regression Network (TURN) model. There are two salient aspects of TURN: (1) TURN jointly predicts action proposals and refines the temporal boundaries by temporal coordinate regression; (2) Fast computation is enabled by unit feature reuse: a long untrimmed video is decomposed into video units, which are reused as basic building blocks of temporal proposals. TURN outperforms the state-of-the-art methods under average recall (AR) by a large margin on THUMOS-14 and ActivityNet datasets, and runs at over 880 frames per second (FPS) on a TITAN X GPU. We further apply TURN as a proposal generation stage for existing temporal action localization pipelines, it outperforms state-of-the-art performance on THUMOS-14 and ActivityNet.

arXiv:1708.02188

PowerAI DDL

As deep neural networks become more complex and input datasets grow larger, it can take days or even weeks to train a deep neural network to the desired accuracy. Therefore, distributed Deep Learning at a massive scale is a critical capability, since it offers the potential to reduce the training time from weeks to hours. In this paper, we present a software-hardware co-optimized distributed Deep Learning system that can achieve near-linear scaling up to hundreds of GPUs. The core algorithm is a multi-ring communication pattern that provides a good tradeoff between latency and bandwidth and adapts to a variety of system configurations. The communication algorithm is implemented as a library for easy use. This library has been integrated into Tensorflow, Caffe, and Torch. We train Resnet-101 on Imagenet 22K with 64 IBM Power8 S822LC servers (256 GPUs) in about 7 hours to an accuracy of 33.8 % validation accuracy. Microsoft's ADAM and Google's DistBelief results did not reach 30 % validation accuracy for Imagenet 22K. Compared to Facebook AI Research's recent paper on 256 GPU training, we use a different communication algorithm, and our combined software and hardware system offers better communication overhead for Resnet-50. A PowerAI DDL enabled version of Torch completed 90 epochs of training on Resnet 50 for 1K classes in 50 minutes using 64 IBM Power8 S822LC servers (256 GPUs).

arXiv:1708.01936

Face Parsing via Recurrent Propagation

Face parsing is an important problem in computer vision that finds numerous applications including recognition and editing. Recently, deep convolutional neural networks (CNNs) have been applied to image parsing and segmentation with the state-of-the-art performance. In this paper, we propose a face parsing algorithm that combines hierarchical representations learned by a CNN, and accurate label propagations achieved by a spatially variant recurrent neural network (RNN). The RNN-based propagation approach enables efficient inference over a global space with the guidance of semantic edges generated by a local convolutional model. Since the convolutional architecture can be shallow and the spatial RNN can have few parameters, the framework is much faster and more light-weighted than the state-of-the-art CNNs for the same task. We apply the proposed model to coarse-grained and fine-grained face parsing. For fine-grained face parsing, we develop a two-stage approach by first identifying the main regions and then segmenting the detail components, which achieves better performance in terms of accuracy and efficiency. With a single GPU, the proposed algorithm parses face images accurately at 300 frames per second, which facilitates real-time applications.

arXiv:1708.02898

An evaluation of large-scale methods for image instance and class discovery

This paper aims at discovering meaningful subsets of related images from large image collections without annotations. We search groups of images related at different levels of semantic, i.e., either instances or visual classes. While k-means is usually considered as the gold standard for this task, we evaluate and show the interest of diffusion methods that have been neglected by the state of the art, such as the Markov Clustering algorithm. We report results on the ImageNet and the Paris500k instance dataset, both enlarged with images from YFCC100M. We evaluate our methods with a labelling cost that reflects how much effort a human would require to correct the generated clusters. Our analysis highlights several properties. First, when powered with an efficient GPU implementation, the cost of the discovery process is small compared to computing the image descriptors, even for collections as large as 100 million images. Second, we show that descriptions selected for instance search improve the discovery of object classes. Third, the Markov Clustering technique consistently outperforms other methods; to our knowledge it has never been considered in this large scale scenario.

arXiv:1708.02711

Tips and Tricks for Visual Question Answering: Learnings from the 2017 Challenge

This paper presents a state-of-the-art model for visual question answering (VQA), which won the first place in the 2017 VQA Challenge. VQA is a task of significant importance for research in artificial intelligence, given its multimodal nature, clear evaluation protocol, and potential real-world applications. The performance of deep neural networks for VQA is very dependent on choices of architectures and hyperparameters. To help further research in the area, we describe in detail our high-performing, though relatively simple model. Through a massive exploration of architectures and hyperparameters representing more than 3,000 GPU-hours, we identified tips and tricks that lead to its success, namely: sigmoid outputs, soft training targets, image features from bottom-up attention, gated tanh activations, output embeddings initialized using GloVe and Google Images, large mini-batches, and smart shuffling of training data. We provide a detailed analysis of their impact on performance to assist others in making an appropriate selection.

arXiv:1703.02529공식 코드 · Papers with Code

NoScope: Optimizing Neural Network Queries over Video at Scale

Recent advances in computer vision-in the form of deep neural networks-have made it possible to query increasing volumes of video data with high accuracy. However, neural network inference is computationally expensive at scale: applying a state-of-the-art object detector in real time (i.e., 30+ frames per second) to a single video requires a $4000 GPU. In response, we present NoScope, a system for querying videos that can reduce the cost of neural network video analysis by up to three orders of magnitude via inference-optimized model search. Given a target video, object to detect, and reference neural network, NoScope automatically searches for and trains a sequence, or cascade, of models that preserves the accuracy of the reference network but is specialized to the target video and are therefore far less computationally expensive. NoScope cascades two types of models: specialized models that forego the full generality of the reference model but faithfully mimic its behavior for the target video and object; and difference detectors that highlight temporal differences across frames. We show that the optimal cascade architecture differs across videos and objects, so NoScope uses an efficient cost-based optimizer to search across models and cascades. With this approach, NoScope achieves two to three order of magnitude speed-ups (265-15,500x real-time) on binary classification tasks over fixed-angle webcam and surveillance video while maintaining accuracy within 1-5% of state-of-the-art neural networks.

arXiv:1612.01501

BrainFrame: A node-level heterogeneous accelerator platform for neuron simulations

Objective: The advent of High-Performance Computing (HPC) in recent years has led to its increasing use in brain study through computational models. The scale and complexity of such models are constantly increasing, leading to challenging computational requirements. Even though modern HPC platforms can often deal with such challenges, the vast diversity of the modeling field does not permit for a single acceleration (or homogeneous) platform to effectively address the complete array of modeling requirements. Approach: In this paper we propose and build BrainFrame, a heterogeneous acceleration platform, incorporating three distinct acceleration technologies, a Dataflow Engine, a Xeon Phi and a GP-GPU. The PyNN framework is also integrated into the platform. As a challenging proof of concept, we analyze the performance of BrainFrame on different instances of a state-of-the-art neuron model, modeling the Inferior- Olivary Nucleus using a biophysically-meaningful, extended Hodgkin-Huxley representation. The model instances take into account not only the neuronal- network dimensions but also different network-connectivity circumstances that can drastically change application workload characteristics. Main results: The synthetic approach of three HPC technologies demonstrated that BrainFrame is better able to cope with the modeling diversity encountered. Our performance analysis shows clearly that the model directly affect performance and all three technologies are required to cope with all the model use cases.

arXiv:1708.05604

Accelerating recurrent neural network training using sequence bucketing and multi-GPU data parallelization

An efficient algorithm for recurrent neural network training is presented. The approach increases the training speed for tasks where a length of the input sequence may vary significantly. The proposed approach is based on the optimal batch bucketing by input sequence length and data parallelization on multiple graphical processing units. The baseline training performance without sequence bucketing is compared with the proposed solution for a different number of buckets. An example is given for the online handwriting recognition task using an LSTM recurrent neural network. The evaluation is performed in terms of the wall clock time, number of epochs, and validation loss value.

arXiv:1708.05997

A Batch Noise Contrastive Estimation Approach for Training Large Vocabulary Language Models

Training large vocabulary Neural Network Language Models (NNLMs) is a difficult task due to the explicit requirement of the output layer normalization, which typically involves the evaluation of the full softmax function over the complete vocabulary. This paper proposes a Batch Noise Contrastive Estimation (B-NCE) approach to alleviate this problem. This is achieved by reducing the vocabulary, at each time step, to the target words in the batch and then replacing the softmax by the noise contrastive estimation approach, where these words play the role of targets and noise samples at the same time. In doing so, the proposed approach can be fully formulated and implemented using optimal dense matrix operations. Applying B-NCE to train different NNLMs on the Large Text Compression Benchmark (LTCB) and the One Billion Word Benchmark (OBWB) shows a significant reduction of the training time with no noticeable degradation of the models performance. This paper also presents a new baseline comparative study of different standard NNLMs on the large OBWB on a single Titan-X GPU.

arXiv:1708.08670

Performance Analysis of Open Source Machine Learning Frameworks for Various Parameters in Single-Threaded and Multi-Threaded Modes

The basic features of some of the most versatile and popular open source frameworks for machine learning (TensorFlow, Deep Learning4j, and H2O) are considered and compared. Their comparative analysis was performed and conclusions were made as to the advantages and disadvantages of these platforms. The performance tests for the de facto standard MNIST data set were carried out on H2O framework for deep learning algorithms designed for CPU and GPU platforms for single-threaded and multithreaded modes of operation Also, we present the results of testing neural networks architectures on H2O platform for various activation functions, stopping metrics, and other parameters of machine learning algorithm. It was demonstrated for the use case of MNIST database of handwritten digits in single-threaded mode that blind selection of these parameters can hugely increase (by 2-3 orders) the runtime without the significant increase of precision. This result can have crucial influence for optimization of available and new machine learning methods, especially for image recognition problems.

arXiv:1708.08180

An Optimized Union-Find Algorithm for Connected Components Labeling Using GPUs

In this paper, we report an optimized union-find (UF) algorithm that can label the connected components on a 2D image efficiently by employing the GPU architecture. The proposed method contains three phases: UF-based local merge, boundary analysis, and link. The coarse labeling in local merge reduces the number atomic operations, while the boundary analysis only manages the pixels on the boundary of each block. Evaluation results showed that the proposed algorithm speed up the average running time by more than 1.3X.

arXiv:1709.00947

Learning Word Embeddings from the Portuguese Twitter Stream: A Study of some Practical Aspects

This paper describes a preliminary study for producing and distributing a large-scale database of embeddings from the Portuguese Twitter stream. We start by experimenting with a relatively small sample and focusing on three challenges: volume of training data, vocabulary size and intrinsic evaluation metrics. Using a single GPU, we were able to scale up vocabulary size from 2048 words embedded and 500K training examples to 32768 words over 10M training examples while keeping a stable validation loss and approximately linear trend on training time per epoch. We also observed that using less than 50\% of the available training examples for each vocabulary size might result in overfitting. Results on intrinsic evaluation show promising performance for a vocabulary size of 32768 words. Nevertheless, intrinsic evaluation metrics suffer from over-sensitivity to their corresponding cosine similarity thresholds, indicating that a wider range of metrics need to be developed to track progress.

arXiv:1609.09850

Latent fingerprint minutia extraction using fully convolutional network

Minutiae play a major role in fingerprint identification. Extracting reliable minutiae is difficult for latent fingerprints which are usually of poor quality. As the limitation of traditional handcrafted features, a fully convolutional network (FCN) is utilized to learn features directly from data to overcome complex background noises. Raw fingerprints are mapped to a correspondingly-sized minutia-score map with a fixed stride. And thus a large number of minutiae will be extracted through a given threshold. Then small regions centering at these minutia points are entered into a convolutional neural network (CNN) to reclassify these minutiae and calculate their orientations. The CNN shares convolutional layers with the fully convolutional network to speed up. 0.45 second is used on average to detect one fingerprint on a GPU. On the NIST SD27 database, we achieve 53\% recall rate and 53\% precise rate that outperform many other algorithms. Our trained model is also visualized to show that we have successfully extracted features preserving ridge information of a latent fingerprint.

arXiv:1709.03329

weedNet: Dense Semantic Weed Classification Using Multispectral Images and MAV for Smart Farming

Selective weed treatment is a critical step in autonomous crop management as related to crop health and yield. However, a key challenge is reliable, and accurate weed detection to minimize damage to surrounding plants. In this paper, we present an approach for dense semantic weed classification with multispectral images collected by a micro aerial vehicle (MAV). We use the recently developed encoder-decoder cascaded Convolutional Neural Network (CNN), Segnet, that infers dense semantic classes while allowing any number of input image channels and class balancing with our sugar beet and weed datasets. To obtain training datasets, we established an experimental field with varying herbicide levels resulting in field plots containing only either crop or weed, enabling us to use the Normalized Difference Vegetation Index (NDVI) as a distinguishable feature for automatic ground truth generation. We train 6 models with different numbers of input channels and condition (fine-tune) it to achieve about 0.8 F1-score and 0.78 Area Under the Curve (AUC) classification metrics. For model deployment, an embedded GPU system (Jetson TX2) is tested for MAV integration. Dataset used in this paper is released to support the community and future work.

arXiv:1707.00424

Parle: parallelizing stochastic gradient descent

We propose a new algorithm called Parle for parallel training of deep networks that converges 2-4x faster than a data-parallel implementation of SGD, while achieving significantly improved error rates that are nearly state-of-the-art on several benchmarks including CIFAR-10 and CIFAR-100, without introducing any additional hyper-parameters. We exploit the phenomenon of flat minima that has been shown to lead to improved generalization error for deep networks. Parle requires very infrequent communication with the parameter server and instead performs more computation on each client, which makes it well-suited to both single-machine, multi-GPU settings and distributed implementations.