PyTorchKR Papers

최신 논문

총 782편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2209.12843

Impact of temporal resolution on convolutional recurrent networks for audio tagging and sound event detection

Many state-of-the-art systems for audio tagging and sound event detection employ convolutional recurrent neural architectures. Typically, they are trained in a mean teacher setting to deal with the heterogeneous annotation of the available data. In this work, we present a thorough analysis of how changing the temporal resolution of these convolutional recurrent neural networks - which can be done by simply adapting their pooling operations - impacts their performance. By using a variety of evaluation metrics, we investigate the effects of adapting this design parameter under several sound recognition scenarios involving different needs in terms of temporal localization.

arXiv:2209.12826

Multi-encoder attention-based architectures for sound recognition with partial visual assistance

Large-scale sound recognition data sets typically consist of acoustic recordings obtained from multimedia libraries. As a consequence, modalities other than audio can often be exploited to improve the outputs of models designed for associated tasks. Frequently, however, not all contents are available for all samples of such a collection: For example, the original material may have been removed from the source platform at some point, and therefore, non-auditory features can no longer be acquired. We demonstrate that a multi-encoder framework can be employed to deal with this issue by applying this method to attention-based deep learning systems, which are currently part of the state of the art in the domain of sound recognition. More specifically, we show that the proposed model extension can successfully be utilized to incorporate partially available visual information into the operational procedures of such networks, which normally only use auditory features during training and inference. Experimentally, we verify that the considered approach leads to improved predictions in a number of evaluation scenarios pertaining to audio tagging and sound event detection. Additionally, we scrutinize some properties and limitations of the presented technique.

arXiv:2209.12148공식 코드 · Papers with Code

Self-Supervised Masked Convolutional Transformer Block for Anomaly Detection

Anomaly detection has recently gained increasing attention in the field of computer vision, likely due to its broad set of applications ranging from product fault detection on industrial production lines and impending event detection in video surveillance to finding lesions in medical scans. Regardless of the domain, anomaly detection is typically framed as a one-class classification task, where the learning is conducted on normal examples only. An entire family of successful anomaly detection methods is based on learning to reconstruct masked normal inputs (e.g. patches, future frames, etc.) and exerting the magnitude of the reconstruction error as an indicator for the abnormality level. Unlike other reconstruction-based methods, we present a novel self-supervised masked convolutional transformer block (SSMCTB) that comprises the reconstruction-based functionality at a core architectural level. The proposed self-supervised block is extremely flexible, enabling information masking at any layer of a neural network and being compatible with a wide range of neural architectures. In this work, we extend our previous self-supervised predictive convolutional attentive block (SSPCAB) with a 3D masked convolutional layer, a transformer for channel-wise attention, as well as a novel self-supervised objective based on Huber loss. Furthermore, we show that our block is applicable to a wider variety of tasks, adding anomaly detection in medical images and thermal videos to the previously considered tasks based on RGB images and surveillance videos. We exhibit the generality and flexibility of SSMCTB by integrating it into multiple state-of-the-art neural models for anomaly detection, bringing forth empirical results that confirm considerable performance improvements on five benchmarks. We release our code and data as open source at: https://github.com/ristea/ssmctb.

arXiv:2209.11007공식 코드 · Papers with Code

Avoiding Post-Processing with Event-Based Detection in Biomedical Signals

Objective: Finding events of interest is a common task in biomedical signal processing. The detection of epileptic seizures and signal artefacts are two key examples. Epoch-based classification is the typical machine learning framework to detect such signal events because of the straightforward application of classical machine learning techniques. Usually, post-processing is required to achieve good performance and enforce temporal dependencies. Designing the right post-processing scheme to convert these classification outputs into events is a tedious, and labor-intensive element of this framework. Methods: We propose an event-based modeling framework that directly works with events as learning targets, stepping away from ad-hoc post-processing schemes to turn model outputs into events. We illustrate the practical power of this framework on simulated data and real-world data, comparing it to epoch-based modeling approaches. Results: We show that event-based modeling (without post-processing) performs on par with or better than epoch-based modeling with extensive post-processing. Conclusion: These results show the power of treating events as direct learning targets, instead of using ad-hoc post-processing to obtain them, severely reducing design effort. Significance: The event-based modeling framework can easily be applied to other event detection problems in signal processing, removing the need for intensive task-specific post-processing.

arXiv:2209.05900

Binaural Signal Representations for Joint Sound Event Detection and Acoustic Scene Classification

Sound event detection (SED) and Acoustic scene classification (ASC) are two widely researched audio tasks that constitute an important part of research on acoustic scene analysis. Considering shared information between sound events and acoustic scenes, performing both tasks jointly is a natural part of a complex machine listening system. In this paper, we investigate the usefulness of several spatial audio features in training a joint deep neural network (DNN) model performing SED and ASC. Experiments are performed for two different datasets containing binaural recordings and synchronous sound event and acoustic scene labels to analyse the differences between performing SED and ASC separately or jointly. The presented results show that the use of specific binaural features, mainly the Generalized Cross Correlation with Phase Transform (GCC-phat) and sines and cosines of phase differences, result in a better performing model in both separate and joint tasks as compared with baseline methods based on logmel energies only.

arXiv:2209.01979

Few-shot Incremental Event Detection

Event detection tasks can enable the quick detection of events from texts and provide powerful support for downstream natural language processing tasks. Most such methods can only detect a fixed set of predefined event classes. To extend them to detect a new class without losing the ability to detect old classes requires costly retraining of the model from scratch. Incremental learning can effectively solve this problem, but it requires abundant data of new classes. In practice, however, the lack of high-quality labeled data of new event classes makes it difficult to obtain enough data for model training. To address the above mentioned issues, we define a new task, few-shot incremental event detection, which focuses on learning to detect a new event class with limited data, while retaining the ability to detect old classes to the extent possible. We created a benchmark dataset IFSED for the few-shot incremental event detection task based on FewEvent and propose two benchmarks, IFSED-K and IFSED-KP. Experimental results show that our approach has a higher F1-score than baseline methods and is more stable.

arXiv:2209.01802공식 코드 · Papers with Code

Sound Event Localization and Detection for Real Spatial Sound Scenes: Event-Independent Network and Data Augmentation Chains

Sound event localization and detection (SELD) is a joint task of sound event detection and direction-of-arrival estimation. In DCASE 2022 Task 3, types of data transform from computationally generated spatial recordings to recordings of real-sound scenes. Our system submitted to the DCASE 2022 Task 3 is based on our previous proposed Event-Independent Network V2 (EINV2) with a novel data augmentation method. Our method employs EINV2 with a track-wise output format, permutation-invariant training, and a soft parameter-sharing strategy, to detect different sound events of the same class but in different locations. The Conformer structure is used for extending EINV2 to learn local and global features. A data augmentation method, which contains several data augmentation chains composed of stochastic combinations of several different data augmentation operations, is utilized to generalize the model. To mitigate the lack of real-scene recordings in the development dataset and the presence of sound events being unbalanced, we exploit FSD50K, AudioSet, and TAU Spatial Room Impulse Response Database (TAU-SRIR DB) to generate simulated datasets for training. We present results on the validation set of Sony-TAu Realistic Spatial Soundscapes 2022 (STARSS22) in detail. Experimental results indicate that the ability to generalize to different environments and unbalanced performance among different classes are two main challenges. We evaluate our proposed method in Task 3 of the DCASE 2022 challenge and obtain the second rank in the teams ranking. Source code is released.

arXiv:2209.02339공식 코드 · Papers with Code

TransCAB: Transferable Clean-Annotation Backdoor to Object Detection with Natural Trigger in Real-World

Object detection is the foundation of various critical computer-vision tasks such as segmentation, object tracking, and event detection. To train an object detector with satisfactory accuracy, a large amount of data is required. However, due to the intensive workforce involved with annotating large datasets, such a data curation task is often outsourced to a third party or relied on volunteers. This work reveals severe vulnerabilities of such data curation pipeline. We propose MACAB that crafts clean-annotated images to stealthily implant the backdoor into the object detectors trained on them even when the data curator can manually audit the images. We observe that the backdoor effect of both misclassification and the cloaking are robustly achieved in the wild when the backdoor is activated with inconspicuously natural physical triggers. Backdooring non-classification object detection with clean-annotation is challenging compared to backdooring existing image classification tasks with clean-label, owing to the complexity of having multiple objects within each frame, including victim and non-victim objects. The efficacy of the MACAB is ensured by constructively i abusing the image-scaling function used by the deep learning framework, ii incorporating the proposed adversarial clean image replica technique, and iii combining poison data selection criteria given constrained attacking budget. Extensive experiments demonstrate that MACAB exhibits more than 90% attack success rate under various real-world scenes. This includes both cloaking and misclassification backdoor effect even restricted with a small attack budget. The poisoned samples cannot be effectively identified by state-of-the-art detection techniques.The comprehensive video demo is at https://youtu.be/MA7L_LpXkp4, which is based on a poison rate of 0.14% for YOLOv4 cloaking backdoor and Faster R-CNN misclassification backdoor.

arXiv:2208.13101

An event detection technique using social media data

People post information about different topics which are in their active vocabulary over social media platforms (like Twitter, Facebook, PInterest and Google+). They follow each other and it is more likely that the person who posts information about current happenings will receive better response. Manual analysis of huge amount of data on social media platforms is difficult. This has opened new research directions for automatic analysis of usercontributed social media documents. Automatic social media data analysis is difficult due to abundant information shared by users. Many researchers use Twitter data for Social Media Analysis (SMA) as the Twitter data is freely available in the public domain. One of the most this research work. Event Detection from social media data is used for different applications like traffic congestion detection, disaster and emergency management, and live news detection. Nature of the information which is shared on twitter platform is short-text, noisy, and ambiguous. Thus, event detection and extraction of event phrases from user-generated and illformed data becomes challenging. To address these challenges, events are extracted from streaming social media data in the form of keyphrases using different cognitive properties. The motivation behind this research work is to provide substantial improvements in the lexical variation of event phrases while detecting events and sub-events from twitter data. In this research work, the approach towards event detection from social media data is divided into three phases namely: Identifying sub-graphs in Microblog Word Co-occurrence Network (WCN) which provides important information about keyphrases; Identifying multiple events from social media data; and Ranking contextual information of event phrases.

arXiv:2208.11467공식 코드 · Papers with Code

Tracking by weakly-supervised learning and graph optimization for whole-embryo C. elegans lineages

Tracking all nuclei of an embryo in noisy and dense fluorescence microscopy data is a challenging task. We build upon a recent method for nuclei tracking that combines weakly-supervised learning from a small set of nuclei center point annotations with an integer linear program (ILP) for optimal cell lineage extraction. Our work specifically addresses the following challenging properties of C. elegans embryo recordings: (1) Many cell divisions as compared to benchmark recordings of other organisms, and (2) the presence of polar bodies that are easily mistaken as cell nuclei. To cope with (1), we devise and incorporate a learnt cell division detector. To cope with (2), we employ a learnt polar body detector. We further propose automated ILP weights tuning via a structured SVM, alleviating the need for tedious manual set-up of a respective grid search. Our method outperforms the previous leader of the cell tracking challenge on the Fluo-N3DH-CE embryo dataset. We report a further extensive quantitative evaluation on two more C. elegans datasets. We will make these datasets public to serve as an extended benchmark for future method development. Our results suggest considerable improvements yielded by our method, especially in terms of the correctness of division event detection and the number and length of fully correct track segments. Code: https://github.com/funkelab/linajea

arXiv:2208.10638

Scalable Hybrid Classification-Regression Solution for High-Frequency Nonintrusive Load Monitoring

Residential buildings with the ability to monitor and control their net-load (sum of load and generation) can provide valuable flexibility to power grid operators. We present a novel multiclass nonintrusive load monitoring (NILM) approach that enables effective net-load monitoring capabilities at high-frequency with minimal additional equipment and cost. The proposed machine learning based solution provides accurate multiclass state predictions while operating at a faster timescale (able to provide a prediction for each 60-Hz ac cycle used in US power grid) without relying on event-detection techniques. We also introduce an innovative hybrid classification-regression method that allows for the prediction of not only load on/off states via classification but also individual load operating power levels via regression. A test bed with eight residential appliances is used for validating the NILM approach. Results show that the overall method has high accuracy and, good scaling and generalization properties. Furthermore, the method is shown to have sufficient response time (within 160ms, corresponding to 10 ac cycles) to support building grid-interactive control at fast timescales relevant to the provision of grid frequency support services.

arXiv:2208.09659

Trigger-free Event Detection via Derangement Reading Comprehension

Event detection (ED), aiming to detect events from texts and categorize them, is vital to understanding actual happenings in real life. However, mainstream event detection models require high-quality expert human annotations of triggers, which are often costly and thus deter the application of ED to new domains. Therefore, in this paper, we focus on low-resource ED without triggers and aim to tackle the following formidable challenges: multi-label classification, insufficient clues, and imbalanced events distribution. We propose a novel trigger-free ED method via Derangement mechanism on a machine Reading Comprehension (DRC) framework. More specifically, we treat the input text as Context and concatenate it with all event type tokens that are deemed as Answers with an omitted default question. So we can leverage the self-attention in pre-trained language models to absorb semantic relations between input text and the event types. Moreover, we design a simple yet effective event derangement module (EDM) to prevent major events from being excessively learned so as to yield a more balanced training process. The experiment results show that our proposed trigger-free ED model is remarkably competitive to mainstream trigger-based models, showing its strong performance on low-source event detection.

arXiv:2208.07969

A Sensor-Based Simulation Method for Spatiotemporal Event Detection

Human movements in urban areas are essential to understand human-environment interactions. However, activities and associated movements are full of uncertainties due to the complexity of a city. In this paper, we propose a novel sensor-based approach for spatiotemporal event detection based on the Discrete Empirical Interpolation Method. Specifically, we first identify the key locations, defined as 'sensors' , which have the strongest correlation with the whole dataset. We then simulate a regular uneventful scenario with the observation data points from those key lo-cations. By comparing the simulated and observation scenarios, events are extracted both spatially and temporally. We apply this method in New York City with taxi trip record data. Results show that this method is effective in detecting when and where events occur.

arXiv:2208.04637

Fisher Matrix Based Fault Detection for PMUs Data in Power Grids

Abnormal event detection is critical in the safe operation of power system. In this paper, using the data collected from phasor measurement units (PMUs), two methods based on Fisher random matrix are proposed to detect faults in power grids. Firstly, the fault detection matrix is constructed and the event detection problem is reformatted as a two-sample covariance matrices test problem. Secondly, the central limit theorem for the linear spectral statistic of the Fisher matrix is derived and a test statistic for testing faults is proposed. To save computing resources, the screening step of fault interval based on the test statistic is designed to check the existence of faults. Then two point-by-point methods are proposed to determine the time of the fault in the interval. One method detects faults by checking whether the largest sample eigenvalue falls outside the supporting set of limiting spectral distribution of the standard Fisher matrix, which can detect the faults with higher accuracy. The other method tests the faults based on the statistic proposed, which has a faster detection speed. Compared with existing works, the simulation results illustrate that two methods proposed in this paper cost less computational time and provide a higher degree of accuracy.

arXiv:2207.14024공식 코드 · Papers with Code

Safety-Enhanced Autonomous Driving Using Interpretable Sensor Fusion Transformer

Large-scale deployment of autonomous vehicles has been continually delayed due to safety concerns. On the one hand, comprehensive scene understanding is indispensable, a lack of which would result in vulnerability to rare but complex traffic situations, such as the sudden emergence of unknown objects. However, reasoning from a global context requires access to sensors of multiple types and adequate fusion of multi-modal sensor signals, which is difficult to achieve. On the other hand, the lack of interpretability in learning models also hampers the safety with unverifiable failure causes. In this paper, we propose a safety-enhanced autonomous driving framework, named Interpretable Sensor Fusion Transformer(InterFuser), to fully process and fuse information from multi-modal multi-view sensors for achieving comprehensive scene understanding and adversarial event detection. Besides, intermediate interpretable features are generated from our framework, which provide more semantics and are exploited to better constrain actions to be within the safe sets. We conducted extensive experiments on CARLA benchmarks, where our model outperforms prior methods, ranking the first on the public CARLA Leaderboard. Our code will be made available at https://github.com/opendilab/InterFuser

arXiv:2207.07935공식 코드 · Papers with Code

Visually-aware Acoustic Event Detection using Heterogeneous Graphs

Perception of auditory events is inherently multimodal relying on both audio and visual cues. A large number of existing multimodal approaches process each modality using modality-specific models and then fuse the embeddings to encode the joint information. In contrast, we employ heterogeneous graphs to explicitly capture the spatial and temporal relationships between the modalities and represent detailed information about the underlying signal. Using heterogeneous graph approaches to address the task of visually-aware acoustic event classification, which serves as a compact, efficient and scalable way to represent data in the form of graphs. Through heterogeneous graphs, we show efficiently modelling of intra- and inter-modality relationships both at spatial and temporal scales. Our model can easily be adapted to different scales of events through relevant hyperparameters. Experiments on AudioSet, a large benchmark, shows that our model achieves state-of-the-art performance.

arXiv:2207.07911공식 코드 · Papers with Code

Few-shot bioacoustic event detection at the DCASE 2022 challenge

Few-shot sound event detection is the task of detecting sound events, despite having only a few labelled examples of the class of interest. This framework is particularly useful in bioacoustics, where often there is a need to annotate very long recordings but the expert annotator time is limited. This paper presents an overview of the second edition of the few-shot bioacoustic sound event detection task included in the DCASE 2022 challenge. A detailed description of the task objectives, dataset, and baselines is presented, together with the main results obtained and characteristics of the submitted systems. This task received submissions from 15 different teams from which 13 scored higher than the baselines. The highest F-score was of 60% on the evaluation set, which leads to a huge improvement over last year's edition. Highly-performing methods made use of prototypical networks, transductive learning, and addressed the variable length of events from all target classes. Furthermore, by analysing results on each of the subsets we can identify the main difficulties that the systems face, and conclude that few-show bioacoustic sound event detection remains an open challenge.

arXiv:2207.08625공식 코드 · Papers with Code

Unifying Event Detection and Captioning as Sequence Generation via Pre-Training

Dense video captioning aims to generate corresponding text descriptions for a series of events in the untrimmed video, which can be divided into two sub-tasks, event detection and event captioning. Unlike previous works that tackle the two sub-tasks separately, recent works have focused on enhancing the inter-task association between the two sub-tasks. However, designing inter-task interactions for event detection and captioning is not trivial due to the large differences in their task specific solutions. Besides, previous event detection methods normally ignore temporal dependencies between events, leading to event redundancy or inconsistency problems. To tackle above the two defects, in this paper, we define event detection as a sequence generation task and propose a unified pre-training and fine-tuning framework to naturally enhance the inter-task association between event detection and captioning. Since the model predicts each event with previous events as context, the inter-dependency between events is fully exploited and thus our model can detect more diverse and consistent events in the video. Experiments on the ActivityNet dataset show that our model outperforms the state-of-the-art methods, and can be further boosted when pre-trained on extra large-scale video-text data. Code is available at \url{https://github.com/QiQAng/UEDVC}.

arXiv:2207.07773공식 코드 · Papers with Code

Segment-level Metric Learning for Few-shot Bioacoustic Event Detection

Few-shot bioacoustic event detection is a task that detects the occurrence time of a novel sound given a few examples. Previous methods employ metric learning to build a latent space with the labeled part of different sound classes, also known as positive events. In this study, we propose a segment-level few-shot learning framework that utilizes both the positive and negative events during model optimization. Training with negative events, which are larger in volume than positive events, can increase the generalization ability of the model. In addition, we use transductive inference on the validation set during training for better adaptation to novel classes. We conduct ablation studies on our proposed method with different setups on input features, training data, and hyper-parameters. Our final system achieves an F-measure of 62.73 on the DCASE 2022 challenge task 5 (DCASE2022-T5) validation set, outperforming the performance of the baseline prototypical network 34.02 by a large margin. Using the proposed method, our submitted system ranks 2nd in DCASE2022-T5. The code of this paper is fully open-sourced at https://github.com/haoheliu/DCASE_2022_Task_5.

arXiv:2207.05204

An Online Data-Driven Method for Microgrid Secondary Voltage and Frequency Control with Ensemble Koopman Modeling

Low inertia, nonlinearity and a high level of uncertainty (varying topologies and operating conditions) pose challenges to microgrid (MG) systemwide operation. This paper proposes an online adaptive Koopman operator optimal control (AKOOC) method for MG secondary voltage and frequency control. Unlike typical data-driven methods that are data-hungry and lack guaranteed stability, the proposed AKOOC requires no warm-up training yet with guaranteed bounded-input-bounded-output (BIBO) stability and even asymptotical stability under some mild conditions. The proposed AKOOC is developed based on an ensemble Koopman state space modeling with full basis functions that combines both linear and nonlinear bases without the need of event detection or switching. An iterative learning method is also developed to exploit model parameters, ensuring the effectiveness and the adaptiveness of the designed control. Simulation studies in the 4-bus (with detailed inner-loop control) MG system and the 34-bus MG system showed improved modeling accuracy and control, verifying the effectiveness of the proposed method subject to various changes of operating conditions even with time delay, measurement noise, and missing measurements.