Pub Date : 2022-03-18DOI: 10.48550/arXiv.2203.10009
D. Avola, Marco Cascio, L. Cinque, Alessio Fagioli, G. Foresti, Marco Raoul Marini, D. Pannone
Nowadays, machine and deep learning techniques are widely used in different areas, ranging from economics to biology. In general, these techniques can be used in two ways: trying to adapt well-known models and architectures to the available data, or designing custom architectures. In both cases, to speed up the research process, it is useful to know which type of models work best for a specific problem and/or data type. By focusing on EEG signal analysis, and for the first time in literature, in this paper a benchmark of machine and deep learning for EEG signal classification is proposed. For our experiments we used the four most widespread models, i.e., multilayer perceptron, convolutional neural network, long short-term memory, and gated recurrent unit, highlighting which one can be a good starting point for developing EEG classification models.
{"title":"Analyzing EEG Data with Machine and Deep Learning: A Benchmark","authors":"D. Avola, Marco Cascio, L. Cinque, Alessio Fagioli, G. Foresti, Marco Raoul Marini, D. Pannone","doi":"10.48550/arXiv.2203.10009","DOIUrl":"https://doi.org/10.48550/arXiv.2203.10009","url":null,"abstract":"Nowadays, machine and deep learning techniques are widely used in different areas, ranging from economics to biology. In general, these techniques can be used in two ways: trying to adapt well-known models and architectures to the available data, or designing custom architectures. In both cases, to speed up the research process, it is useful to know which type of models work best for a specific problem and/or data type. By focusing on EEG signal analysis, and for the first time in literature, in this paper a benchmark of machine and deep learning for EEG signal classification is proposed. For our experiments we used the four most widespread models, i.e., multilayer perceptron, convolutional neural network, long short-term memory, and gated recurrent unit, highlighting which one can be a good starting point for developing EEG classification models.","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"56 1","pages":"335-345"},"PeriodicalIF":0.0,"publicationDate":"2022-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79165950","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-16DOI: 10.48550/arXiv.2203.08688
Alex Falcon, G. Serra, O. Lanz
Due to the amount of videos and related captions uploaded every hour, deep learning-based solutions for cross-modal video retrieval are attracting more and more attention. A typical approach consists in learning a joint text-video embedding space, where the similarity of a video and its associated caption is maximized, whereas a lower similarity is enforced with all the other captions, called negatives. This approach assumes that only the video and caption pairs in the dataset are valid, but different captions positives may also describe its visual contents, hence some of them may be wrongly penalized. To address this shortcoming, we propose the Relevance-Aware Negatives and Positives mining (RANP) which, based on the semantics of the negatives, improves their selection while also increasing the similarity of other valid positives. We explore the influence of these techniques on two videotext datasets: EPIC-Kitchens-100 and MSR-VTT. By using the proposed techniques, we achieve considerable improvements in terms of nDCG and mAP, leading to state-of-the-art results, e.g. +5.3% nDCG and +3.0% mAP on EPIC-Kitchens-100. We share code and pretrained models at https://github.com/aranciokov/ranp.
{"title":"Learning video retrieval models with relevance-aware online mining","authors":"Alex Falcon, G. Serra, O. Lanz","doi":"10.48550/arXiv.2203.08688","DOIUrl":"https://doi.org/10.48550/arXiv.2203.08688","url":null,"abstract":"Due to the amount of videos and related captions uploaded every hour, deep learning-based solutions for cross-modal video retrieval are attracting more and more attention. A typical approach consists in learning a joint text-video embedding space, where the similarity of a video and its associated caption is maximized, whereas a lower similarity is enforced with all the other captions, called negatives. This approach assumes that only the video and caption pairs in the dataset are valid, but different captions positives may also describe its visual contents, hence some of them may be wrongly penalized. To address this shortcoming, we propose the Relevance-Aware Negatives and Positives mining (RANP) which, based on the semantics of the negatives, improves their selection while also increasing the similarity of other valid positives. We explore the influence of these techniques on two videotext datasets: EPIC-Kitchens-100 and MSR-VTT. By using the proposed techniques, we achieve considerable improvements in terms of nDCG and mAP, leading to state-of-the-art results, e.g. +5.3% nDCG and +3.0% mAP on EPIC-Kitchens-100. We share code and pretrained models at https://github.com/aranciokov/ranp.","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"26 1","pages":"182-194"},"PeriodicalIF":0.0,"publicationDate":"2022-03-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87366829","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-15DOI: 10.48550/arXiv.2203.07973
Donato Cafarelli, Luca Ciampi, Lucia Vadicamo, C. Gennaro, A. Berton, M. Paterni, C. Benvenuti, M. Passera, F. Falchi
Modern Unmanned Aerial Vehicles (UAV) equipped with cameras can play an essential role in speeding up the identification and rescue of people who have fallen overboard, i.e., man overboard (MOB). To this end, Artificial Intelligence techniques can be leveraged for the automatic understanding of visual data acquired from drones. However, detecting people at sea in aerial imagery is challenging primarily due to the lack of specialized annotated datasets for training and testing detectors for this task. To fill this gap, we introduce and publicly release the MOBDrone benchmark, a collection of more than 125K drone-view images in a marine environment under several conditions, such as different altitudes, camera shooting angles, and illumination. We manually annotated more than 180K objects, of which about 113K man overboard, precisely localizing them with bounding boxes. Moreover, we conduct a thorough performance analysis of several state-of-the-art object detectors on the MOBDrone data, serving as baselines for further research.
{"title":"MOBDrone: a Drone Video Dataset for Man OverBoard Rescue","authors":"Donato Cafarelli, Luca Ciampi, Lucia Vadicamo, C. Gennaro, A. Berton, M. Paterni, C. Benvenuti, M. Passera, F. Falchi","doi":"10.48550/arXiv.2203.07973","DOIUrl":"https://doi.org/10.48550/arXiv.2203.07973","url":null,"abstract":"Modern Unmanned Aerial Vehicles (UAV) equipped with cameras can play an essential role in speeding up the identification and rescue of people who have fallen overboard, i.e., man overboard (MOB). To this end, Artificial Intelligence techniques can be leveraged for the automatic understanding of visual data acquired from drones. However, detecting people at sea in aerial imagery is challenging primarily due to the lack of specialized annotated datasets for training and testing detectors for this task. To fill this gap, we introduce and publicly release the MOBDrone benchmark, a collection of more than 125K drone-view images in a marine environment under several conditions, such as different altitudes, camera shooting angles, and illumination. We manually annotated more than 180K objects, of which about 113K man overboard, precisely localizing them with bounding boxes. Moreover, we conduct a thorough performance analysis of several state-of-the-art object detectors on the MOBDrone data, serving as baselines for further research.","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"49 1","pages":"633-644"},"PeriodicalIF":0.0,"publicationDate":"2022-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73565925","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-13DOI: 10.1007/978-3-031-06433-3_21
David Freire-Obregón, J. Lorenzo-Navarro, M. C. Santana
{"title":"Decontextualized I3D ConvNet for ultra-distance runners performance analysis at a glance","authors":"David Freire-Obregón, J. Lorenzo-Navarro, M. C. Santana","doi":"10.1007/978-3-031-06433-3_21","DOIUrl":"https://doi.org/10.1007/978-3-031-06433-3_21","url":null,"abstract":"","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"62 1","pages":"242-253"},"PeriodicalIF":0.0,"publicationDate":"2022-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79055060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-03-11DOI: 10.48550/arXiv.2203.05807
Nathan Hubens, M. Mancas, B. Gosselin, Marius Preda, T. Zaharia
Neural network pruning is a widely used strategy for reducing model storage and computing requirements. It allows to lower the complexity of the network by introducing sparsity in the weights. Because taking advantage of sparse matrices is still challenging, pruning is often performed in a structured way, i.e. removing entire convolution filters in the case of ConvNets, according to a chosen pruning criteria. Common pruning criteria, such as l1-norm or movement, usually do not consider the individual utility of filters, which may lead to: (1) the removal of filters exhibiting rare, thus important and discriminative behaviour, and (2) the retaining of filters with redundant information. In this paper, we present a technique solving those two issues, and which can be appended to any pruning criteria. This technique ensures that the criteria of selection focuses on redundant filters, while retaining the rare ones, thus maximizing the variety of remaining filters. The experimental results, carried out on different datasets (CIFAR-10, CIFAR-100 and CALTECH-101) and using different architectures (VGG-16 and ResNet-18) demonstrate that it is possible to achieve similar sparsity levels while maintaining a higher performance when appending our filter selection technique to pruning criteria. Moreover, we assess the quality of the found sparse sub-networks by applying the Lottery Ticket Hypothesis and find that the addition of our method allows to discover better performing tickets in most cases
{"title":"Improve Convolutional Neural Network Pruning by Maximizing Filter Variety","authors":"Nathan Hubens, M. Mancas, B. Gosselin, Marius Preda, T. Zaharia","doi":"10.48550/arXiv.2203.05807","DOIUrl":"https://doi.org/10.48550/arXiv.2203.05807","url":null,"abstract":"Neural network pruning is a widely used strategy for reducing model storage and computing requirements. It allows to lower the complexity of the network by introducing sparsity in the weights. Because taking advantage of sparse matrices is still challenging, pruning is often performed in a structured way, i.e. removing entire convolution filters in the case of ConvNets, according to a chosen pruning criteria. Common pruning criteria, such as l1-norm or movement, usually do not consider the individual utility of filters, which may lead to: (1) the removal of filters exhibiting rare, thus important and discriminative behaviour, and (2) the retaining of filters with redundant information. In this paper, we present a technique solving those two issues, and which can be appended to any pruning criteria. This technique ensures that the criteria of selection focuses on redundant filters, while retaining the rare ones, thus maximizing the variety of remaining filters. The experimental results, carried out on different datasets (CIFAR-10, CIFAR-100 and CALTECH-101) and using different architectures (VGG-16 and ResNet-18) demonstrate that it is possible to achieve similar sparsity levels while maintaining a higher performance when appending our filter selection technique to pruning criteria. Moreover, we assess the quality of the found sparse sub-networks by applying the Lottery Ticket Hypothesis and find that the addition of our method allows to discover better performing tickets in most cases","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"73 1","pages":"379-390"},"PeriodicalIF":0.0,"publicationDate":"2022-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85743937","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-02-28DOI: 10.1007/978-3-031-06427-2_44
Nicolo Lucchesi, Antonio Carta, Vincenzo Lomonaco
{"title":"Avalanche RL: a Continual Reinforcement Learning Library","authors":"Nicolo Lucchesi, Antonio Carta, Vincenzo Lomonaco","doi":"10.1007/978-3-031-06427-2_44","DOIUrl":"https://doi.org/10.1007/978-3-031-06427-2_44","url":null,"abstract":"","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"4 1","pages":"524-535"},"PeriodicalIF":0.0,"publicationDate":"2022-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76401821","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-02-04DOI: 10.1007/978-3-031-06430-2_6
Cristina Mata, Nick Locascio, Mohammed Azeem Sheikh, Kenny Kihara, Daniel L. Fischetti
{"title":"StandardSim: A Synthetic Dataset For Retail Environments","authors":"Cristina Mata, Nick Locascio, Mohammed Azeem Sheikh, Kenny Kihara, Daniel L. Fischetti","doi":"10.1007/978-3-031-06430-2_6","DOIUrl":"https://doi.org/10.1007/978-3-031-06430-2_6","url":null,"abstract":"","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"468 1","pages":"65-76"},"PeriodicalIF":0.0,"publicationDate":"2022-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72776501","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-24DOI: 10.1007/978-3-031-06430-2_38
Valerio Paolicelli, A. Tavera, Gabriele Berton, C. Masone, Barbara Caputo
{"title":"Learning Semantics for Visual Place Recognition through Multi-Scale Attention","authors":"Valerio Paolicelli, A. Tavera, Gabriele Berton, C. Masone, Barbara Caputo","doi":"10.1007/978-3-031-06430-2_38","DOIUrl":"https://doi.org/10.1007/978-3-031-06430-2_38","url":null,"abstract":"","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"51 1","pages":"454-466"},"PeriodicalIF":0.0,"publicationDate":"2022-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76544587","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-06430-2_12
Antonio Greco, A. Saggese, Bruno Vento
{"title":"A Robust and Efficient Overhead People Counting System for Retail Applications","authors":"Antonio Greco, A. Saggese, Bruno Vento","doi":"10.1007/978-3-031-06430-2_12","DOIUrl":"https://doi.org/10.1007/978-3-031-06430-2_12","url":null,"abstract":"","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"15 1","pages":"139-150"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75604669","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-01-01DOI: 10.1007/978-3-031-06430-2_3
M. Zohaib, M. Taiana, Milind G. Padalkar, A. D. Bue
{"title":"3D Key-Points Estimation from Single-View RGB Images","authors":"M. Zohaib, M. Taiana, Milind G. Padalkar, A. D. Bue","doi":"10.1007/978-3-031-06430-2_3","DOIUrl":"https://doi.org/10.1007/978-3-031-06430-2_3","url":null,"abstract":"","PeriodicalId":74527,"journal":{"name":"Proceedings of the ... International Conference on Image Analysis and Processing. International Conference on Image Analysis and Processing","volume":"30 1","pages":"27-38"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74354866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：