PyTorchKR Papers

최신 논문

총 5,576편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2211.12222

Event Transformer+. A multi-purpose solution for efficient event data processing

Event cameras record sparse illumination changes with high temporal resolution and high dynamic range. Thanks to their sparse recording and low consumption, they are increasingly used in applications such as AR/VR and autonomous driving. Current top-performing methods often ignore specific event-data properties, leading to the development of generic but computationally expensive algorithms, while event-aware methods do not perform as well. We propose Event Transformer+, that improves our seminal work evtprev EvT with a refined patch-based event representation and a more robust backbone to achieve more accurate results, while still benefiting from event-data sparsity to increase its efficiency. Additionally, we show how our system can work with different data modalities and propose specific output heads, for event-stream predictions (i.e. action recognition) and per-pixel predictions (dense depth estimation). Evaluation results show better performance to the state-of-the-art while requiring minimal computation resources, both on GPU and CPU.

arXiv:2211.11222공식 코드 · Papers with Code

Embedding a Differentiable Mel-cepstral Synthesis Filter to a Neural Speech Synthesis System

This paper integrates a classic mel-cepstral synthesis filter into a modern neural speech synthesis system towards end-to-end controllable speech synthesis. Since the mel-cepstral synthesis filter is explicitly embedded in neural waveform models in the proposed system, both voice characteristics and the pitch of synthesized speech are highly controlled via a frequency warping parameter and fundamental frequency, respectively. We implement the mel-cepstral synthesis filter as a differentiable and GPU-friendly module to enable the acoustic and waveform models in the proposed system to be simultaneously optimized in an end-to-end manner. Experiments show that the proposed system improves speech quality from a baseline system maintaining controllability. The core PyTorch modules used in the experiments will be publicly available on GitHub.

arXiv:2211.10551

A Practical Stereo Depth System for Smart Glasses

We present the design of a productionized end-to-end stereo depth sensing system that does pre-processing, online stereo rectification, and stereo depth estimation with a fallback to monocular depth estimation when rectification is unreliable. The output of our depth sensing system is then used in a novel view generation pipeline to create 3D computational photography effects using point-of-view images captured by smart glasses. All these steps are executed on-device on the stringent compute budget of a mobile phone, and because we expect the users can use a wide range of smartphones, our design needs to be general and cannot be dependent on a particular hardware or ML accelerator such as a smartphone GPU. Although each of these steps is well studied, a description of a practical system is still lacking. For such a system, all these steps need to work in tandem with one another and fallback gracefully on failures within the system or less than ideal input data. We show how we handle unforeseen changes to calibration, e.g., due to heat, robustly support depth estimation in the wild, and still abide by the memory and latency constraints required for a smooth user experience. We show that our trained models are fast, and run in less than 1s on a six-year-old Samsung Galaxy S8 phone's CPU. Our models generalize well to unseen data and achieve good results on Middlebury and in-the-wild images captured from the smart glasses.

arXiv:2211.11131공식 코드 · Papers with Code

Doubly Contrastive End-to-End Semantic Segmentation for Autonomous Driving under Adverse Weather

Road scene understanding tasks have recently become crucial for self-driving vehicles. In particular, real-time semantic segmentation is indispensable for intelligent self-driving agents to recognize roadside objects in the driving area. As prior research works have primarily sought to improve the segmentation performance with computationally heavy operations, they require far significant hardware resources for both training and deployment, and thus are not suitable for real-time applications. As such, we propose a doubly contrastive approach to improve the performance of a more practical lightweight model for self-driving, specifically under adverse weather conditions such as fog, nighttime, rain and snow. Our proposed approach exploits both image- and pixel-level contrasts in an end-to-end supervised learning scheme without requiring a memory bank for global consistency or the pretraining step used in conventional contrastive methods. We validate the effectiveness of our method using SwiftNet on the ACDC dataset, where it achieves up to 1.34%p improvement in mIoU (ResNet-18 backbone) at 66.7 FPS (2048x1024 resolution) on a single RTX 3080 Mobile GPU at inference. Furthermore, we demonstrate that replacing image-level supervision with self-supervision achieves comparable performance when pre-trained with clear weather images.

arXiv:2211.11018

MagicVideo: Efficient Video Generation With Latent Diffusion Models

We present an efficient text-to-video generation framework based on latent diffusion models, termed MagicVideo. MagicVideo can generate smooth video clips that are concordant with the given text descriptions. Due to a novel and efficient 3D U-Net design and modeling video distributions in a low-dimensional space, MagicVideo can synthesize video clips with 256x256 spatial resolution on a single GPU card, which takes around 64x fewer computations than the Video Diffusion Models (VDM) in terms of FLOPs. In specific, unlike existing works that directly train video models in the RGB space, we use a pre-trained VAE to map video clips into a low-dimensional latent space and learn the distribution of videos' latent codes via a diffusion model. Besides, we introduce two new designs to adapt the U-Net denoiser trained on image tasks to video data: a frame-wise lightweight adaptor for the image-to-video distribution adjustment and a directed temporal attention module to capture temporal dependencies across frames. Thus, we can exploit the informative weights of convolution operators from a text-to-image model for accelerating video training. To ameliorate the pixel dithering in the generated videos, we also propose a novel VideoVAE auto-encoder for better RGB reconstruction. We conduct extensive experiments and demonstrate that MagicVideo can generate high-quality video clips with either realistic or imaginary content. Refer to \url{https://magicvideo.github.io/#} for more examples.

arXiv:2211.10981

Real-time Local Feature with Global Visual Information Enhancement

Local feature provides compact and invariant image representation for various visual tasks. Current deep learning-based local feature algorithms always utilize convolution neural network (CNN) architecture with limited receptive field. Besides, even with high-performance GPU devices, the computational efficiency of local features cannot be satisfactory. In this paper, we tackle such problems by proposing a CNN-based local feature algorithm. The proposed method introduces a global enhancement module to fuse global visual clues in a light-weight network, and then optimizes the network by novel deep reinforcement learning scheme from the perspective of local feature matching task. Experiments on the public benchmarks demonstrate that the proposal can achieve considerable robustness against visual interference and meanwhile run in real time.

arXiv:2211.11316

EHSNet: End-to-End Holistic Learning Network for Large-Size Remote Sensing Image Semantic Segmentation

This paper presents EHSNet, a new end-to-end segmentation network designed for the holistic learning of large-size remote sensing image semantic segmentation (LRISS). Large-size remote sensing images (LRIs) can lead to GPU memory exhaustion due to their extremely large size, which has been handled in previous works through either global-local fusion or multi-stage refinement, both of which are limited in their ability to fully exploit the abundant information available in LRIs. Unlike them, EHSNet features three memory-friendly modules to utilize the characteristics of LRIs: a long-range dependency module to develop long-range spatial context, an efficient cross-correlation module to build holistic contextual relationships, and a boundary-aware enhancement module to preserve complete object boundaries. Moreover, EHSNet manages to process holistic LRISS with the aid of memory offloading. To the best of our knowledge, EHSNet is the first method capable of performing holistic LRISS. To make matters better, EHSNet outperforms previous state-of-the-art competitors by a significant margin of +5.65 mIoU on FBP and +4.28 mIoU on Inria Aerial, demonstrating its effectiveness. We hope that EHSNet will provide a new perspective for LRISS. The code and models will be made publicly available.

arXiv:2211.11647

Benchmarking Edge Computing Devices for Grape Bunches and Trunks Detection using Accelerated Object Detection Single Shot MultiBox Deep Learning Models

Purpose: Visual perception enables robots to perceive the environment. Visual data is processed using computer vision algorithms that are usually time-expensive and require powerful devices to process the visual data in real-time, which is unfeasible for open-field robots with limited energy. This work benchmarks the performance of different heterogeneous platforms for object detection in real-time. This research benchmarks three architectures: embedded GPU -- Graphical Processing Units (such as NVIDIA Jetson Nano 2 GB and 4 GB, and NVIDIA Jetson TX2), TPU -- Tensor Processing Unit (such as Coral Dev Board TPU), and DPU -- Deep Learning Processor Unit (such as in AMD-Xilinx ZCU104 Development Board, and AMD-Xilinx Kria KV260 Starter Kit). Method: The authors used the RetinaNet ResNet-50 fine-tuned using the natural VineSet dataset. After the trained model was converted and compiled for target-specific hardware formats to improve the execution efficiency. Conclusions and Results: The platforms were assessed in terms of performance of the evaluation metrics and efficiency (time of inference). Graphical Processing Units (GPUs) were the slowest devices, running at 3 FPS to 5 FPS, and Field Programmable Gate Arrays (FPGAs) were the fastest devices, running at 14 FPS to 25 FPS. The efficiency of the Tensor Processing Unit (TPU) is irrelevant and similar to NVIDIA Jetson TX2. TPU and GPU are the most power-efficient, consuming about 5W. The performance differences, in the evaluation metrics, across devices are irrelevant and have an F1 of about 70 % and mean Average Precision (mAP) of about 60 %.

arXiv:2211.10655

Solving 3D Inverse Problems using Pre-trained 2D Diffusion Models

Diffusion models have emerged as the new state-of-the-art generative model with high quality samples, with intriguing properties such as mode coverage and high flexibility. They have also been shown to be effective inverse problem solvers, acting as the prior of the distribution, while the information of the forward model can be granted at the sampling stage. Nonetheless, as the generative process remains in the same high dimensional (i.e. identical to data dimension) space, the models have not been extended to 3D inverse problems due to the extremely high memory and computational cost. In this paper, we combine the ideas from the conventional model-based iterative reconstruction with the modern diffusion models, which leads to a highly effective method for solving 3D medical image reconstruction tasks such as sparse-view tomography, limited angle tomography, compressed sensing MRI from pre-trained 2D diffusion models. In essence, we propose to augment the 2D diffusion prior with a model-based prior in the remaining direction at test time, such that one can achieve coherent reconstructions across all dimensions. Our method can be run in a single commodity GPU, and establishes the new state-of-the-art, showing that the proposed method can perform reconstructions of high fidelity and accuracy even in the most extreme cases (e.g. 2-view 3D tomography). We further reveal that the generalization capacity of the proposed method is surprisingly high, and can be used to reconstruct volumes that are entirely different from the training dataset.

arXiv:2211.10586공식 코드 · Papers with Code

Scaling Up Dataset Distillation to ImageNet-1K with Constant Memory

Dataset Distillation is a newly emerging area that aims to distill large datasets into much smaller and highly informative synthetic ones to accelerate training and reduce storage. Among various dataset distillation methods, trajectory-matching-based methods (MTT) have achieved SOTA performance in many tasks, e.g., on CIFAR-10/100. However, due to exorbitant memory consumption when unrolling optimization through SGD steps, MTT fails to scale to large-scale datasets such as ImageNet-1K. Can we scale this SOTA method to ImageNet-1K and does its effectiveness on CIFAR transfer to ImageNet-1K? To answer these questions, we first propose a procedure to exactly compute the unrolled gradient with constant memory complexity, which allows us to scale MTT to ImageNet-1K seamlessly with ~6x reduction in memory footprint. We further discover that it is challenging for MTT to handle datasets with a large number of classes, and propose a novel soft label assignment that drastically improves its convergence. The resulting algorithm sets new SOTA on ImageNet-1K: we can scale up to 50 IPCs (Image Per Class) on ImageNet-1K on a single GPU (all previous methods can only scale to 2 IPCs on ImageNet-1K), leading to the best accuracy (only 5.9% accuracy drop against full dataset training) while utilizing only 4.2% of the number of data points - an 18.2% absolute gain over prior SOTA. Our code is available at https://github.com/justincui03/tesla

arXiv:2211.11446

SMAUG: Sparse Masked Autoencoder for Efficient Video-Language Pre-training

Video-language pre-training is crucial for learning powerful multi-modal representation. However, it typically requires a massive amount of computation. In this paper, we develop SMAUG, an efficient pre-training framework for video-language models. The foundation component in SMAUG is masked autoencoders. Different from prior works which only mask textual inputs, our masking strategy considers both visual and textual modalities, providing a better cross-modal alignment and saving more pre-training costs. On top of that, we introduce a space-time token sparsification module, which leverages context information to further select only "important" spatial regions and temporal frames for pre-training. Coupling all these designs allows our method to enjoy both competitive performances on text-to-video retrieval and video question answering tasks, and much less pre-training costs by 1.9X or more. For example, our SMAUG only needs about 50 NVIDIA A6000 GPU hours for pre-training to attain competitive performances on these two video-language tasks across six popular benchmarks.

arXiv:2211.10841공식 코드 · Papers with Code

SeDR: Segment Representation Learning for Long Documents Dense Retrieval

Recently, Dense Retrieval (DR) has become a promising solution to document retrieval, where document representations are used to perform effective and efficient semantic search. However, DR remains challenging on long documents, due to the quadratic complexity of its Transformer-based encoder and the finite capacity of a low-dimension embedding. Current DR models use suboptimal strategies such as truncating or splitting-and-pooling to long documents leading to poor utilization of whole document information. In this work, to tackle this problem, we propose Segment representation learning for long documents Dense Retrieval (SeDR). In SeDR, Segment-Interaction Transformer is proposed to encode long documents into document-aware and segment-sensitive representations, while it holds the complexity of splitting-and-pooling and outperforms other segment-interaction patterns on DR. Since GPU memory requirements for long document encoding causes insufficient negatives for DR training, Late-Cache Negative is further proposed to provide additional cache negatives for optimizing representation learning. Experiments on MS MARCO and TREC-DL datasets show that SeDR achieves superior performance among DR models, and confirm the effectiveness of SeDR on long document retrieval.

arXiv:2211.09359

How to Fine-Tune Vision Models with SGD

SGD and AdamW are the two most used optimizers for fine-tuning large neural networks in computer vision. When the two methods perform the same, SGD is preferable because it uses less memory (12 bytes/parameter with momentum and 8 bytes/parameter without) than AdamW (16 bytes/parameter). However, on a suite of downstream tasks, especially those with distribution shifts, we find that fine-tuning with AdamW performs substantially better than SGD on modern Vision Transformer and ConvNeXt models. We find that large gaps in performance between SGD and AdamW occur when the fine-tuning gradients in the first "embedding" layer are much larger than in the rest of the model. Our analysis suggests an easy fix that works consistently across datasets and models: freezing the embedding layer (less than 1% of the parameters) leads to SGD with or without momentum performing slightly better than AdamW while using less memory (e.g., on ViT-L, SGD uses 33% less GPU memory). Our insights result in state-of-the-art accuracies on five popular distribution shift benchmarks: WILDS-FMoW, WILDS-Camelyon, BREEDS-Living-17, Waterbirds, and DomainNet.

arXiv:2211.08842

Fast and Accurate FSA System Using ELBERT: An Efficient and Lightweight BERT

With the development of deep learning and Transformer-based pre-trained models like BERT, the accuracy of many NLP tasks has been dramatically improved. However, the large number of parameters and computations also pose challenges for their deployment. For instance, using BERT can improve the predictions in the financial sentiment analysis (FSA) task but slow it down, where speed and accuracy are equally important in terms of profits. To address these issues, we first propose an efficient and lightweight BERT (ELBERT) along with a novel confidence-window-based (CWB) early exit mechanism. Based on ELBERT, an innovative method to accelerate text processing on the GPU platform is developed, solving the difficult problem of making the early exit mechanism work more effectively with a large input batch size. Afterward, a fast and high-accuracy FSA system is built. Experimental results show that the proposed CWB early exit mechanism achieves significantly higher accuracy than existing early exit methods on BERT under the same computation cost. By using this acceleration method, our FSA system can boost the processing speed by nearly 40 times to over 1000 texts per second with sufficient accuracy, which is nearly twice as fast as FastBERT, thus providing a more powerful text processing capability for modern trading systems.

arXiv:2211.08506공식 코드 · Papers with Code

ParticleGrid: Enabling Deep Learning using 3D Representation of Materials

From AlexNet to Inception, autoencoders to diffusion models, the development of novel and powerful deep learning models and learning algorithms has proceeded at breakneck speeds. In part, we believe that rapid iteration of model architecture and learning techniques by a large community of researchers over a common representation of the underlying entities has resulted in transferable deep learning knowledge. As a result, model scale, accuracy, fidelity, and compute performance have dramatically increased in computer vision and natural language processing. On the other hand, the lack of a common representation for chemical structure has hampered similar progress. To enable transferable deep learning, we identify the need for a robust 3-dimensional representation of materials such as molecules and crystals. The goal is to enable both materials property prediction and materials generation with 3D structures. While computationally costly, such representations can model a large set of chemical structures. We propose $\textit{ParticleGrid}$, a SIMD-optimized library for 3D structures, that is designed for deep learning applications and to seamlessly integrate with deep learning frameworks. Our highly optimized grid generation allows for generating grids on the fly on the CPU, reducing storage and GPU compute and memory requirements. We show the efficacy of 3D grids generated via $\textit{ParticleGrid}$ and accurately predict molecular energy properties using a 3D convolutional neural network. Our model is able to get 0.006 mean square error and nearly match the values calculated using computationally costly density functional theory at a fraction of the time.

arXiv:2211.08666공식 코드 · Papers with Code

Revisiting Training-free NAS Metrics: An Efficient Training-based Method

Recent neural architecture search (NAS) works proposed training-free metrics to rank networks which largely reduced the search cost in NAS. In this paper, we revisit these training-free metrics and find that: (1) the number of parameters (\#Param), which is the most straightforward training-free metric, is overlooked in previous works but is surprisingly effective, (2) recent training-free metrics largely rely on the \#Param information to rank networks. Our experiments show that the performance of recent training-free metrics drops dramatically when the \#Param information is not available. Motivated by these observations, we argue that metrics less correlated with the \#Param are desired to provide additional information for NAS. We propose a light-weight training-based metric which has a weak correlation with the \#Param while achieving better performance than training-free metrics at a lower search cost. Specifically, on DARTS search space, our method completes searching directly on ImageNet in only 2.6 GPU hours and achieves a top-1/top-5 error rate of 24.1\%/7.1\%, which is competitive among state-of-the-art NAS methods. Codes are available at \url{https://github.com/taoyang1122/Revisit_TrainingFree_NAS}

arXiv:2211.08144

Monocular BEV Perception of Road Scenes via Front-to-Top View Projection

HD map reconstruction is crucial for autonomous driving. LiDAR-based methods are limited due to expensive sensors and time-consuming computation. Camera-based methods usually need to perform road segmentation and view transformation separately, which often causes distortion and missing content. To push the limits of the technology, we present a novel framework that reconstructs a local map formed by road layout and vehicle occupancy in the bird's-eye view given a front-view monocular image only. We propose a front-to-top view projection (FTVP) module, which takes the constraint of cycle consistency between views into account and makes full use of their correlation to strengthen the view transformation and scene understanding. In addition, we also apply multi-scale FTVP modules to propagate the rich spatial information of low-level features to mitigate spatial deviation of the predicted object location. Experiments on public benchmarks show that our method achieves the state-of-the-art performance in the tasks of road layout estimation, vehicle occupancy estimation, and multi-class semantic estimation. For multi-class semantic estimation, in particular, our model outperforms all competitors by a large margin. Furthermore, our model runs at 25 FPS on a single GPU, which is efficient and applicable for real-time panorama HD map reconstruction.

arXiv:2211.06769

Realistic Bokeh Effect Rendering on Mobile GPUs, Mobile AI & AIM 2022 challenge: Report

As mobile cameras with compact optics are unable to produce a strong bokeh effect, lots of interest is now devoted to deep learning-based solutions for this task. In this Mobile AI challenge, the target was to develop an efficient end-to-end AI-based bokeh effect rendering approach that can run on modern smartphone GPUs using TensorFlow Lite. The participants were provided with a large-scale EBB! bokeh dataset consisting of 5K shallow / wide depth-of-field image pairs captured using the Canon 7D DSLR camera. The runtime of the resulting models was evaluated on the Kirin 9000's Mali GPU that provides excellent acceleration results for the majority of common deep learning ops. A detailed description of all models developed in this challenge is provided in this paper.

arXiv:2211.07429공식 코드 · Papers with Code

Accounting for Temporal Variability in Functional Magnetic Resonance Imaging Improves Prediction of Intelligence

Neuroimaging-based prediction methods for intelligence and cognitive abilities have seen a rapid development in literature. Among different neuroimaging modalities, prediction based on functional connectivity (FC) has shown great promise. Most literature has focused on prediction using static FC, but there are limited investigations on the merits of such analysis compared to prediction based on dynamic FC or region level functional magnetic resonance imaging (fMRI) times series that encode temporal variability. To account for the temporal dynamics in fMRI data, we propose a deep neural network involving bi-directional long short-term memory (bi-LSTM) approach that also incorporates feature selection mechanism. The proposed pipeline is implemented via an efficient GPU computation framework and applied to predict intelligence scores based on region level fMRI time series as well as dynamic FC. We compare the prediction performance for different intelligence measures based on static FC, dynamic FC, and region level time series acquired from the Adolescent Brain Cognitive Development (ABCD) study involving close to 7000 individuals. Our detailed analysis illustrates that static FC consistently has inferior prediction performance compared to region level time series or dynamic FC for unimodal rest and task fMRI experiments, and in almost all cases using a combination of task and rest features. In addition, the proposed bi-LSTM pipeline based on region level time series identifies several shared and differential important brain regions across task and rest fMRI experiments that drive intelligence prediction. A test-retest analysis of the selected features shows strong reliability across cross-validation folds. Given the large sample size from ABCD study, our results provide strong evidence that superior prediction of intelligence can be achieved by accounting for temporal variations in fMRI.

arXiv:2211.06918

Towards a Dynamic Composability Approach for using Heterogeneous Systems in Remote Sensing

Influenced by the advances in data and computing, the scientific practice increasingly involves machine learning and artificial intelligence driven methods which requires specialized capabilities at the system-, science- and service-level in addition to the conventional large-capacity supercomputing approaches. The latest distributed architectures built around the composability of data-centric applications led to the emergence of a new ecosystem for container coordination and integration. However, there is still a divide between the application development pipelines of existing supercomputing environments, and these new dynamic environments that disaggregate fluid resource pools through accessible, portable and re-programmable interfaces. New approaches for dynamic composability of heterogeneous systems are needed to further advance the data-driven scientific practice for the purpose of more efficient computing and usable tools for specific scientific domains. In this paper, we present a novel approach for using composable systems in the intersection between scientific computing, artificial intelligence (AI), and remote sensing domain. We describe the architecture of a first working example of a composable infrastructure that federates Expanse, an NSF-funded supercomputer, with Nautilus, a Kubernetes-based GPU geo-distributed cluster. We also summarize a case study in wildfire modeling, that demonstrates the application of this new infrastructure in scientific workflows: a composed system that bridges the insights from edge sensing, AI and computing capabilities with a physics-driven simulation.