PyTorchKR Papers

최신 논문

총 9,339편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2304.11857

Accurate and Efficient Event-based Semantic Segmentation Using Adaptive Spiking Encoder-Decoder Network

Spiking neural networks (SNNs), known for their low-power, event-driven computation and intrinsic temporal dynamics, are emerging as promising solutions for processing dynamic, asynchronous signals from event-based sensors. Despite their potential, SNNs face challenges in training and architectural design, resulting in limited performance in challenging event-based dense prediction tasks compared to artificial neural networks (ANNs). In this work, we develop an efficient spiking encoder-decoder network (SpikingEDN) for large-scale event-based semantic segmentation tasks. To enhance the learning efficiency from dynamic event streams, we harness the adaptive threshold which improves network accuracy, sparsity and robustness in streaming inference. Moreover, we develop a dual-path Spiking Spatially-Adaptive Modulation module, which is specifically tailored to enhance the representation of sparse events and multi-modal inputs, thereby considerably improving network performance. Our SpikingEDN attains a mean intersection over union (MIoU) of 72.57\% on the DDD17 dataset and 58.32\% on the larger DSEC-Semantic dataset, showing competitive results to the state-of-the-art ANNs while requiring substantially fewer computational resources. Our results shed light on the untapped potential of SNNs in event-based vision applications. The source code will be made publicly available.

arXiv:2304.12141

Variational Diffusion Auto-encoder: Latent Space Extraction from Pre-trained Diffusion Models

As a widely recognized approach to deep generative modeling, Variational Auto-Encoders (VAEs) still face challenges with the quality of generated images, often presenting noticeable blurriness. This issue stems from the unrealistic assumption that approximates the conditional data distribution, $p(\textbf{x} | \textbf{z})$, as an isotropic Gaussian. In this paper, we propose a novel solution to address these issues. We illustrate how one can extract a latent space from a pre-existing diffusion model by optimizing an encoder to maximize the marginal data log-likelihood. Furthermore, we demonstrate that a decoder can be analytically derived post encoder-training, employing the Bayes rule for scores. This leads to a VAE-esque deep latent variable model, which discards the need for Gaussian assumptions on $p(\textbf{x} | \textbf{z})$ or the training of a separate decoder network. Our method, which capitalizes on the strengths of pre-trained diffusion models and equips them with latent spaces, results in a significant enhancement to the performance of VAEs.

arXiv:2304.11705공식 코드 · Papers with Code

Walking Your LiDOG: A Journey Through Multiple Domains for LiDAR Semantic Segmentation

The ability to deploy robots that can operate safely in diverse environments is crucial for developing embodied intelligent agents. As a community, we have made tremendous progress in within-domain LiDAR semantic segmentation. However, do these methods generalize across domains? To answer this question, we design the first experimental setup for studying domain generalization (DG) for LiDAR semantic segmentation (DG-LSS). Our results confirm a significant gap between methods, evaluated in a cross-domain setting: for example, a model trained on the source dataset (SemanticKITTI) obtains $26.53$ mIoU on the target data, compared to $48.49$ mIoU obtained by the model trained on the target domain (nuScenes). To tackle this gap, we propose the first method specifically designed for DG-LSS, which obtains $34.88$ mIoU on the target domain, outperforming all baselines. Our method augments a sparse-convolutional encoder-decoder 3D segmentation network with an additional, dense 2D convolutional decoder that learns to classify a birds-eye view of the point cloud. This simple auxiliary task encourages the 3D network to learn features that are robust to sensor placement shifts and resolution, and are transferable across domains. With this work, we aim to inspire the community to develop and evaluate future models in such cross-domain conditions.

arXiv:2304.11433

Conditional Denoising Diffusion for Sequential Recommendation

Generative models have attracted significant interest due to their ability to handle uncertainty by learning the inherent data distributions. However, two prominent generative models, namely Generative Adversarial Networks (GANs) and Variational AutoEncoders (VAEs), exhibit challenges that impede achieving optimal performance in sequential recommendation tasks. Specifically, GANs suffer from unstable optimization, while VAEs are prone to posterior collapse and over-smoothed generations. The sparse and noisy nature of sequential recommendation further exacerbates these issues. In response to these limitations, we present a conditional denoising diffusion model, which includes a sequence encoder, a cross-attentive denoising decoder, and a step-wise diffuser. This approach streamlines the optimization and generation process by dividing it into easier and tractable steps in a conditional autoregressive manner. Furthermore, we introduce a novel optimization schema that incorporates both cross-divergence loss and contrastive loss. This novel training schema enables the model to generate high-quality sequence/item representations and meanwhile precluding collapse. We conducted comprehensive experiments on four benchmark datasets, and the superior performance achieved by our model attests to its efficacy.

arXiv:2304.11053

A Comparison of Semi-Supervised Learning Techniques for Streaming ASR at Scale

Unpaired text and audio injection have emerged as dominant methods for improving ASR performance in the absence of a large labeled corpus. However, little guidance exists on deploying these methods to improve production ASR systems that are trained on very large supervised corpora and with realistic requirements like a constrained model size and CPU budget, streaming capability, and a rich lattice for rescoring and for downstream NLU tasks. In this work, we compare three state-of-the-art semi-supervised methods encompassing both unpaired text and audio as well as several of their combinations in a controlled setting using joint training. We find that in our setting these methods offer many improvements beyond raw WER, including substantial gains in tail-word WER, decoder computation during inference, and lattice density.

arXiv:2304.10704

Interactive System-wise Anomaly Detection

Anomaly detection, where data instances are discovered containing feature patterns different from the majority, plays a fundamental role in various applications. However, it is challenging for existing methods to handle the scenarios where the instances are systems whose characteristics are not readily observed as data. Appropriate interactions are needed to interact with the systems and identify those with abnormal responses. Detecting system-wise anomalies is a challenging task due to several reasons including: how to formally define the system-wise anomaly detection problem; how to find the effective activation signal for interacting with systems to progressively collect the data and learn the detector; how to guarantee stable training in such a non-stationary scenario with real-time interactions? To address the challenges, we propose InterSAD (Interactive System-wise Anomaly Detection). Specifically, first, we adopt Markov decision process to model the interactive systems, and define anomalous systems as anomalous transition and anomalous reward systems. Then, we develop an end-to-end approach which includes an encoder-decoder module that learns system embeddings, and a policy network to generate effective activation for separating embeddings of normal and anomaly systems. Finally, we design a training method to stabilize the learning process, which includes a replay buffer to store historical interaction data and allow them to be re-sampled. Experiments on two benchmark environments, including identifying the anomalous robotic systems and detecting user data poisoning in recommendation models, demonstrate the superiority of InterSAD compared with state-of-the-art baselines methods.

arXiv:2304.10679

A Plug-and-Play Defensive Perturbation for Copyright Protection of DNN-based Applications

Wide deployment of deep neural networks (DNNs) based applications (e.g., style transfer, cartoonish), stimulating the requirement of copyright protection of such application's production. Although some traditional visible copyright techniques are available, they would introduce undesired traces and result in a poor user experience. In this paper, we propose a novel plug-and-play invisible copyright protection method based on defensive perturbation for DNN-based applications (i.e., style transfer). Rather than apply the perturbation to attack the DNNs model, we explore the potential utilization of perturbation in copyright protection. Specifically, we project the copyright information to the defensive perturbation with the designed copyright encoder, which is added to the image to be protected. Then, we extract the copyright information from the encoded copyrighted image with the devised copyright decoder. Furthermore, we use a robustness module to strengthen the decoding capability of the decoder toward images with various distortions (e.g., JPEG compression), which may be occurred when the user posts the image on social media. To ensure the image quality of encoded images and decoded copyright images, a loss function was elaborately devised. Objective and subjective experiment results demonstrate the effectiveness of the proposed method. We have also conducted physical world tests on social media (i.e., Wechat and Twitter) by posting encoded copyright images. The results show that the copyright information in the encoded image saved from social media can still be correctly extracted.

arXiv:2304.10891

Transformer-based models and hardware acceleration analysis in autonomous driving: A survey

Transformer architectures have exhibited promising performance in various autonomous driving applications in recent years. On the other hand, its dedicated hardware acceleration on portable computational platforms has become the next critical step for practical deployment in real autonomous vehicles. This survey paper provides a comprehensive overview, benchmark, and analysis of Transformer-based models specifically tailored for autonomous driving tasks such as lane detection, segmentation, tracking, planning, and decision-making. We review different architectures for organizing Transformer inputs and outputs, such as encoder-decoder and encoder-only structures, and explore their respective advantages and disadvantages. Furthermore, we discuss Transformer-related operators and their hardware acceleration schemes in depth, taking into account key factors such as quantization and runtime. We specifically illustrate the operator level comparison between layers from convolutional neural network, Swin-Transformer, and Transformer with 4D encoder. The paper also highlights the challenges, trends, and current insights in Transformer-based models, addressing their hardware deployment and acceleration issues within the context of long-term autonomous driving applications.

arXiv:2304.10087

Learning Diffractive Optical Communication Around Arbitrary Opaque Occlusions

Free-space optical systems are emerging for high data rate communication and transfer of information in indoor and outdoor settings. However, free-space optical communication becomes challenging when an occlusion blocks the light path. Here, we demonstrate, for the first time, a direct communication scheme, passing optical information around a fully opaque, arbitrarily shaped obstacle that partially or entirely occludes the transmitter's field-of-view. In this scheme, an electronic neural network encoder and a diffractive optical network decoder are jointly trained using deep learning to transfer the optical information or message of interest around the opaque occlusion of an arbitrary shape. The diffractive decoder comprises successive spatially-engineered passive surfaces that process optical information through light-matter interactions. Following its training, the encoder-decoder pair can communicate any arbitrary optical information around opaque occlusions, where information decoding occurs at the speed of light propagation. For occlusions that change their size and/or shape as a function of time, the encoder neural network can be retrained to successfully communicate with the existing diffractive decoder, without changing the physical layer(s) already deployed. We also validate this framework experimentally in the terahertz spectrum using a 3D-printed diffractive decoder to communicate around a fully opaque occlusion. Scalable for operation in any wavelength regime, this scheme could be particularly useful in emerging high data-rate free-space communication systems.

arXiv:2304.10294

OptoGPT: A Foundation Model for Inverse Design in Optical Multilayer Thin Film Structures

Optical multilayer thin film structures have been widely used in numerous photonic applications. However, existing inverse design methods have many drawbacks because they either fail to quickly adapt to different design targets, or are difficult to suit for different types of structures, e.g., designing for different materials at each layer. These methods also cannot accommodate versatile design situations under different angles and polarizations. In addition, how to benefit practical fabrications and manufacturing has not been extensively considered yet. In this work, we introduce OptoGPT (Opto Generative Pretrained Transformer), a decoder-only transformer, to solve all these drawbacks and issues simultaneously.

arXiv:2304.09764

An End-to-End Vehicle Trajcetory Prediction Framework

Anticipating the motion of neighboring vehicles is crucial for autonomous driving, especially on congested highways where even slight motion variations can result in catastrophic collisions. An accurate prediction of a future trajectory does not just rely on the previous trajectory, but also, more importantly, a simulation of the complex interactions between other vehicles nearby. Most state-of-the-art networks built to tackle the problem assume readily available past trajectory points, hence lacking a full end-to-end pipeline with direct video-to-output mechanism. In this article, we thus propose a novel end-to-end architecture that takes raw video inputs and outputs future trajectory predictions. It first extracts and tracks the 3D location of the nearby vehicles via multi-head attention-based regression networks as well as non-linear optimization. This provides the past trajectory points which then feeds into the trajectory prediction algorithm consisting of an attention-based LSTM encoder-decoder architecture, which allows it to model the complicated interdependence between the vehicles and make an accurate prediction of the future trajectory points of the surrounding vehicles. The proposed model is evaluated on the large-scale BLVD dataset, and has also been implemented on CARLA. The experimental results demonstrate that our approach outperforms various state-of-the-art models.

arXiv:2304.09407공식 코드 · Papers with Code

Pointerformer: Deep Reinforced Multi-Pointer Transformer for the Traveling Salesman Problem

Traveling Salesman Problem (TSP), as a classic routing optimization problem originally arising in the domain of transportation and logistics, has become a critical task in broader domains, such as manufacturing and biology. Recently, Deep Reinforcement Learning (DRL) has been increasingly employed to solve TSP due to its high inference efficiency. Nevertheless, most of existing end-to-end DRL algorithms only perform well on small TSP instances and can hardly generalize to large scale because of the drastically soaring memory consumption and computation time along with the enlarging problem scale. In this paper, we propose a novel end-to-end DRL approach, referred to as Pointerformer, based on multi-pointer Transformer. Particularly, Pointerformer adopts both reversible residual network in the encoder and multi-pointer network in the decoder to effectively contain memory consumption of the encoder-decoder architecture. To further improve the performance of TSP solutions, Pointerformer employs both a feature augmentation method to explore the symmetries of TSP at both training and inference stages as well as an enhanced context embedding approach to include more comprehensive context information in the query. Extensive experiments on a randomly generated benchmark and a public benchmark have shown that, while achieving comparative results on most small-scale TSP instances as SOTA DRL approaches do, Pointerformer can also well generalize to large-scale TSPs.

arXiv:2304.08682공식 코드 · Papers with Code

Learning Situation Hyper-Graphs for Video Question Answering

Answering questions about complex situations in videos requires not only capturing the presence of actors, objects, and their relations but also the evolution of these relationships over time. A situation hyper-graph is a representation that describes situations as scene sub-graphs for video frames and hyper-edges for connected sub-graphs and has been proposed to capture all such information in a compact structured form. In this work, we propose an architecture for Video Question Answering (VQA) that enables answering questions related to video content by predicting situation hyper-graphs, coined Situation Hyper-Graph based Video Question Answering (SHG-VQA). To this end, we train a situation hyper-graph decoder to implicitly identify graph representations with actions and object/human-object relationships from the input video clip. and to use cross-attention between the predicted situation hyper-graphs and the question embedding to predict the correct answer. The proposed method is trained in an end-to-end manner and optimized by a VQA loss with the cross-entropy function and a Hungarian matching loss for the situation graph prediction. The effectiveness of the proposed architecture is extensively evaluated on two challenging benchmarks: AGQA and STAR. Our results show that learning the underlying situation hyper-graphs helps the system to significantly improve its performance for novel challenges of video question-answering tasks.

arXiv:2304.08502

CyFormer: Accurate State-of-Health Prediction of Lithium-Ion Batteries via Cyclic Attention

Predicting the State-of-Health (SoH) of lithium-ion batteries is a fundamental task of battery management systems on electric vehicles. It aims at estimating future SoH based on historical aging data. Most existing deep learning methods rely on filter-based feature extractors (e.g., CNN or Kalman filters) and recurrent time sequence models. Though efficient, they generally ignore cyclic features and the domain gap between training and testing batteries. To address this problem, we present CyFormer, a transformer-based cyclic time sequence model for SoH prediction. Instead of the conventional CNN-RNN structure, we adopt an encoder-decoder architecture. In the encoder, row-wise and column-wise attention blocks effectively capture intra-cycle and inter-cycle connections and extract cyclic features. In the decoder, the SoH queries cross-attend to these features to form the final predictions. We further utilize a transfer learning strategy to narrow the domain gap between the training and testing set. To be specific, we use fine-tuning to shift the model to a target working condition. Finally, we made our model more efficient by pruning. The experiment shows that our method attains an MAE of 0.75\% with only 10\% data for fine-tuning on a testing battery, surpassing prior methods by a large margin. Effective and robust, our method provides a potential solution for all cyclic time sequence prediction tasks.

논문 ID: document/1008653

Deep Double Incomplete Multi-view Multi-label Learning with Incomplete Labels and Missing Views

View missing and label missing are two challenging problems in the applications of multi-view multi-label classification scenery. In the past years, many efforts have been made to address the incomplete multi-view learning or incomplete multi-label learning problem. However, few works can simultaneously handle the challenging case with both two incomplete issues. In this paper, we propose a new incomplete multi-view multi-label learning network to address this challenging issue. The proposed method is composed of four major parts: view-specific deep feature extraction network, weighted representation fusion module, classification module, and view-specific deep decoder network. By respectively integrating the view missing information and label missing information into the weighted fusion module and classification module, the proposed method can effectively reduce the negative influence caused by such two incomplete issues and sufficiently explore the available data and label information to obtain the most discriminative feature extractor and classifier. Furthermore, our method can be trained in both supervised and semi-supervised manners, which has important implications for flexible deployment. Experimental results on five benchmarks in supervised and semi-supervised cases demonstrate that the proposed method can greatly enhance the classification performance on the difficult incomplete multi-view multi-label classification tasks with missing labels and missing views.

arXiv:2304.08431공식 코드 · Papers with Code

Prak: An automatic phonetic alignment tool for Czech

Labeling speech down to the identity and time boundaries of phones is a labor-intensive part of phonetic research. To simplify this work, we created a free open-source tool generating phone sequences from Czech text and time-aligning them with audio. Low architecture complexity makes the design approachable for students of phonetics. Acoustic model ReLU NN with 56k weights was trained using PyTorch on small CommonVoice data. Alignment and variant selection decoder is implemented in Python with matrix library. A Czech pronunciation generator is composed of simple rule-based blocks capturing the logic of the language where possible, allowing modification of transcription approach details. Compared to tools used until now, data preparation efficiency improved, the tool is usable on Mac, Linux and Windows in Praat GUI or command line, achieves mostly correct pronunciation variant choice including glottal stop detection, algorithmically captures most of Czech assimilation logic and is both didactic and practical.

arXiv:2304.08138공식 코드 · Papers with Code

Typos-aware Bottlenecked Pre-Training for Robust Dense Retrieval

Current dense retrievers (DRs) are limited in their ability to effectively process misspelled queries, which constitute a significant portion of query traffic in commercial search engines. The main issue is that the pre-trained language model-based encoders used by DRs are typically trained and fine-tuned using clean, well-curated text data. Misspelled queries are typically not found in the data used for training these models, and thus misspelled queries observed at inference time are out-of-distribution compared to the data used for training and fine-tuning. Previous efforts to address this issue have focused on \textit{fine-tuning} strategies, but their effectiveness on misspelled queries remains lower than that of pipelines that employ separate state-of-the-art spell-checking components. To address this challenge, we propose ToRoDer (TypOs-aware bottlenecked pre-training for RObust DEnse Retrieval), a novel re-training strategy for DRs that increases their robustness to misspelled queries while preserving their effectiveness in downstream retrieval tasks. ToRoDer utilizes an encoder-decoder architecture where the encoder takes misspelled text with masked tokens as input and outputs bottlenecked information to the decoder. The decoder then takes as input the bottlenecked embeddings, along with token embeddings of the original text with the misspelled tokens masked out. The pre-training task is to recover the masked tokens for both the encoder and decoder. Our extensive experimental results and detailed ablation studies show that DRs pre-trained with ToRoDer exhibit significantly higher effectiveness on misspelled queries, sensibly closing the gap with pipelines that use a separate, complex spell-checker component, while retaining their effectiveness on correctly spelled queries.

arXiv:2304.08115공식 코드 · Papers with Code

An Empirical Study of Multitask Learning to Improve Open Domain Dialogue Systems

Autoregressive models used to generate responses in open-domain dialogue systems often struggle to take long-term context into account and to maintain consistency over a dialogue. Previous research in open-domain dialogue generation has shown that the use of \emph{auxiliary tasks} can introduce inductive biases that encourage the model to improve these qualities. However, most previous research has focused on encoder-only or encoder/decoder models, while the use of auxiliary tasks in \emph{decoder-only} autoregressive models is under-explored. This paper describes an investigation where four different auxiliary tasks are added to small and medium-sized GPT-2 models fine-tuned on the PersonaChat and DailyDialog datasets. The results show that the introduction of the new auxiliary tasks leads to small but consistent improvement in evaluations of the investigated models.

arXiv:2304.07957

A Question-Answering Approach to Key Value Pair Extraction from Form-like Document Images

In this paper, we present a new question-answering (QA) based key-value pair extraction approach, called KVPFormer, to robustly extracting key-value relationships between entities from form-like document images. Specifically, KVPFormer first identifies key entities from all entities in an image with a Transformer encoder, then takes these key entities as \textbf{questions} and feeds them into a Transformer decoder to predict their corresponding \textbf{answers} (i.e., value entities) in parallel. To achieve higher answer prediction accuracy, we propose a coarse-to-fine answer prediction approach further, which first extracts multiple answer candidates for each identified question in the coarse stage and then selects the most likely one among these candidates in the fine stage. In this way, the learning difficulty of answer prediction can be effectively reduced so that the prediction accuracy can be improved. Moreover, we introduce a spatial compatibility attention bias into the self-attention/cross-attention mechanism for \Ours{} to better model the spatial interactions between entities. With these new techniques, our proposed \Ours{} achieves state-of-the-art results on FUNSD and XFUND datasets, outperforming the previous best-performing method by 7.2\% and 13.2\% in F1 score, respectively.

arXiv:2304.07611

A CTC Alignment-based Non-autoregressive Transformer for End-to-end Automatic Speech Recognition

Recently, end-to-end models have been widely used in automatic speech recognition (ASR) systems. Two of the most representative approaches are connectionist temporal classification (CTC) and attention-based encoder-decoder (AED) models. Autoregressive transformers, variants of AED, adopt an autoregressive mechanism for token generation and thus are relatively slow during inference. In this paper, we present a comprehensive study of a CTC Alignment-based Single-Step Non-Autoregressive Transformer (CASS-NAT) for end-to-end ASR. In CASS-NAT, word embeddings in the autoregressive transformer (AT) are substituted with token-level acoustic embeddings (TAE) that are extracted from encoder outputs with the acoustical boundary information offered by the CTC alignment. TAE can be obtained in parallel, resulting in a parallel generation of output tokens. During training, Viterbi-alignment is used for TAE generation, and multiple training strategies are further explored to improve the word error rate (WER) performance. During inference, an error-based alignment sampling method is investigated in depth to reduce the alignment mismatch in the training and testing processes. Experimental results show that the CASS-NAT has a WER that is close to AT on various ASR tasks, while providing a ~24x inference speedup. With and without self-supervised learning, we achieve new state-of-the-art results for non-autoregressive models on several datasets. We also analyze the behavior of the CASS-NAT decoder to explain why it can perform similarly to AT. We find that TAEs have similar functionality to word embeddings for grammatical structures, which might indicate the possibility of learning some semantic information from TAEs without a language model.