检索结果-内蒙古大学图书馆

3rd International Conference on Pattern Recognition and Artificial Intelligence, ICPRAI 2022

作者： Savchenko, Andrey V. Laboratory of Algorithms and Technologies for Network Analysis HSE University Nizhny Novgorod Russia

ISBN: (纸本)9783031090363

In this paper, the personalization of the video-based frame-level facial expression recognition is studied for multi-user systems if a small amount of short videos are available for each user. At first, embeddings of each video frame are computed using deep convolutional neural network pre-trained on a large emotional dataset of static images. Next, a dataset of videos is used to train a subject-independent emotion classifier, such as feed-forward neural network or frame attention network. Finally, it is proposed to fine-tune this neural classifier on the videos of each user of interest. As a result, every user is associated with his or her own emotional model. The classifier in a multi-user system is chosen by an appropriate video-based face recognition method. The experimental study with the RAMAS dataset demonstrates the significant (up to 25%) increase in accuracy of the proposed approach when compared to a subject-independent facial expression recognition. © 2022, Springer Nature Switzerland AG.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

Improved Credit Scoring Model with Hyperparameter Optimization 23rd

Improved Credit Scoring Model with Hyperparameter Optimizati...

引用

23rd International Conference on Mathematical Optimization Theory and Operations Research, MOTOR 2024

作者： Marciano, Chiara Guarracino, Mario Rosario Bernhardt, Brian Daniel University of Cassino and Southern Lazio Cassino Italy Laboratory of Algorithms and Technologies for Network Analysis National Research University Higher School of Economics - Nizhny Novgorod Nizhny Novgorod Russia

ISBN: (纸本)9783031733642

This article investigates the application of machine learning techniques for predicting corporate default risk. In the credit scoring domain, the class imbalance problem is prevalent, with defaulted cases typically being much less numerous than non-defaulted ones. To address challenges posed by unbalanced datasets, various calibration methodologies, including data sampling strategies and Synthetic Minority Oversampling Technique (SMOTE), are explored to identify the most effective strategy for the specific case. Additionally, hyperparameter tuning plays a crucial role in enhancing machine learning model performance. Therefore, we employ hyperparameter optimization techniques such as Grid Search, Random Search, Bayesian Optimization, and Adaptive Particle Swarm Optimization. The primary objective is to compare these techniques and determine the optimal strategy for predicting company default events. Our evaluation incorporates standard parameters such as accuracy, balanced accuracy, specificity, sensitivity, and F1 score. Another factor to consider to discriminate the various approaches is the execution time, to be able to develop models that can be implemented in market products. The experimental results reveal which approaches result in significantly improved classification performance. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

关键词： Particle swarm optimization (PSO)

来源：评论

学校读者我要写书评

暂无评论

Effective face recognition based on anomaly image detection and sequential analysis of neural descriptors 9

Effective face recognition based on anomaly image detection ...

引用

9th IEEE International Conference on Information Technology and Nanotechnology, ITNT 2023

作者： Sokolova, Anastasiia Savchenko, Andrey V. Hse University Nizhny Novgorod Russia Sber Ai Lab Moscow Russia Hse University Laboratory of Algorithms and Technologies for Network Analysis Nizhny Novgorod Russia

ISBN: (纸本)9798350397338

In this paper, we explore the possibility to improve efficiency of face recognition using information about anomaly input images. Indeed, modern publicly-available datasets typically contain images of mostly middle-aged and Caucasian people, which cause most algorithms to fail on photos of older people or children, rarer ethnicities, poor-quality images, etc. Detection of such anomaly data and its subsequent rejection helps to improve the classification accuracy. We propose a novel algorithm, at the first stage of which a convolutional neural network is used to detect anomalies in input images. This network is trained on a specially created set of rare data. The second stage is the sequential analysis of neural descriptors extracted from input faces to improve the computational efficiency of classification. An experimental study on the VGGFace2 and MS-Celeb-1M datasets using neural network descriptors, including contemporary InsightFace models, demonstrated the effectiveness of the proposed algorithm. © 2023 IEEE.

关键词： Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

Group-Level Affect Recognition in Video Using Deviation of Frame Features 10th

Group-Level Affect Recognition in Video Using Deviation of...

引用

10th International Conference on analysis of Images, Social networks and Texts, AIST 2021

作者： Savchenko, Andrey V. Savchenko, Lyudmila V. Belova, Natalya S. HSE University Laboratory of Algorithms and Technologies for Network Analysis Nizhny Novgorod Russia HSE University Moscow Russia

ISBN: (纸本)9783031164996

In this paper, we propose the novel video-based group-level emotion recognition algorithm. At first, the faces are detected in each video frame, and their features are extracted using a lightweight neural network, e.g., MobileNet pre-trained on large emotional dataset, such as AffectNet. The frame descriptor is defined as a concatenation of STAT features (max, average, standard deviation, etc.). The descriptor of the whole video is computed as a deviation of the frame descriptors, and the resulting video features are fed into a classifier. Experimental results for the VGAF dataset from the EmotiW 2020 challenge demonstrate that the proposed approach has 1% greater accuracy than the best-known single model. It is also at least 5% better than any other facial processing technique. Moreover, a blending of facial expression recognition with a processing of audio features extracted by the OpenSMILE library is comparable with the best-known ensemble for this dataset. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

Device-Specific Facial Descriptors: Winning a Lottery with a SuperNet 27

Device-Specific Facial Descriptors: Winning a Lottery with a...

引用

27th European Conference on Artificial Intelligence, ECAI 2024

作者： Savchenko, Andrey Maslov, Dmitry Makarov, Ilya Sber AI Lab Moscow Russia Laboratory of Algorithms and Technologies for Network Analysis HSE University Nizhny Novgorod Russia Moscow Russia ISP RAS Research Center for Trusted Artificial Intelligence Moscow Russia

ISBN: (纸本)9781643685489

We address the challenge of devising neural network architectures to extract facial descriptors across diverse mobile and edge *** neural architecture search, we introduce a novel framework that selects optimal subnetworks from a SuperNet using an evolutionary *** a surrogate gradient boosting classifier to avoid direct accuracy estimation of subnetworks on validation sets, our approach swiftly delivers the most efficient and accurate models tailored to specific devices within *** versatility through an Android demo app, our framework excels in tasks like face recognition and emotion understanding across various devices, achieving real-time processing and superior accuracy compared to existing mobile models. © 2024 The Authors.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Summarization of Videos from Online Events Based on Multimodal Emotion Recognition

Summarization of Videos from Online Events Based on Multimod...

引用

2022 International Russian Automation Conference, RusAutoCon 2022

作者： Abdrahimov, Amir Savchenko, Andrey V. Hse University Nizhny Novgorod Russia Hse University Laboratory of Algorithms and Technologies for Network Analysis Nizhny Novgorod Russia

ISBN: (数字)9781665466554

ISBN: (纸本)9781665466554

In this paper, we propose a novel video summarization technique for automatic affect analysis of participants of an online event. At first, face verification neural network is used to cluster facial regions that correspond to each participant. Next, emotional features are extracted from each face by using EfficientNet model obtained in the previous paper of the author. The features of several consecutive frames are combined into a single descriptor that is used to classify emotions. In addition, audio features are extracted using wav2vec, and an ensemble of audio and video classifiers predicts emotions for each face. Finally, dependence of these emotions on time is visualized in special color charts. In the experimental study with the AFEW dataset it was demonstrated that the proposed approach makes it possible to obtain the best-known validation accuracy 67.88%. The models were optimized using OpenVINO and gained reasonable performance even if Nvidia GPUs are unavailable. The models and source code are publicly available at https://***/amirabdrahimov/multimodal-emotion-recognition. © 2022 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

CA-SER: Cross-Attention Feature Fusion for Speech Emotion Recognition 27

CA-SER: Cross-Attention Feature Fusion for Speech Emotion Re...

引用

27th European Conference on Artificial Intelligence, ECAI 2024

作者： Deeb, Bashar M. Savchenko, Andrey Makarov, Ilya MIPT Moscow Russia Laboratory of Algorithms and Technologies for Network Analysis HSE University Nizhny Novgorod Russia Sber AI Lab Moscow Russia Moscow Russia ISP RAS Research Center for Trusted Artificial Intelligence Moscow Russia

ISBN: (纸本)9781643685489

In this paper, we introduce a novel tool for speech emotion recognition, CA-SER, that borrows self-supervised learning to extract semantic speech representations from a pre-trained wav2vec 2.0 model and combine them with spectral audio features to improve speech emotion recognition. Our approach involves a self-attention encoder on MFCC features to capture meaningful patterns in audio sequences. These MFCC features are combined with high-level representations using a multi-head cross-attention mechanism. Evaluation of speech emotion recognition on the IEMOCAP dataset shows that our system achieves a weighted accuracy of 74.6%, outperforming most existing techniques. © 2024 The Authors.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

HSEmotion Team at ABAW-8 Competition: Audiovisual Ambivalence/Hesitancy, Emotional Mimicry Intensity and Facial Expression Recognition

arXiv

引用

arXiv 2025年

作者： Savchenko, Andrey V. HSE University Laboratory of Algorithms and Technologies for Network Analysis Nizhny Novgorod Russia Sber AI Lab Moscow Russia

This article presents our results for the eighth Affective Behavior analysis in-the-Wild (ABAW) competition. We combine facial emotional descriptors extracted by pre-trained models, namely, our EmotiEffLib library, with acoustic features and embeddings of texts recognized from speech. The frame-level features are aggregated and fed into simple classifiers, e.g., multi-layered perceptron (feed-forward neural network with one hidden layer), to predict ambivalence/hesitancy and facial expressions. In the latter case, we also use the pre-trained facial expression recognition model to select high-score video frames and prevent their processing with a domain-specific video classifier. The video-level prediction of emotional mimicry intensity is implemented by simply aggregating frame-level features and training a multi-layered perceptron. Experimental results for three tasks from the ABAW challenge demonstrate that our approach significantly increases validation metrics compared to existing *** Codes 68T10 Copyright © 2025, The Authors. All rights reserved.

关键词： Multilayer neural networks

来源：评论

学校读者我要写书评

暂无评论

HSEmotion Team at the 6th ABAW Competition: Facial Expressions, Valence-Arousal and Emotion Intensity Prediction

arXiv

引用

arXiv 2024年

作者： Savchenko, Andrey V. Sber AI Lab Moscow Russia HSE University Laboratory of Algorithms and Technologies for Network Analysis Nizhny Novgorod Russia

This article presents our results for the sixth Affective Behavior analysis in-the-wild (ABAW) competition. To improve the trustworthiness of facial analysis, we study the possibility of using pre-trained deep models that extract reliable emotional features without the need to fine-tune the neural networks for a downstream task. In particular, we introduce several lightweight models based on MobileViT, MobileFaceNet, EfficientNet, and DDAMFN architectures trained in multi-task scenarios to recognize facial expressions, valence, and arousal on static photos. These neural networks extract frame-level features fed into a simple classifier, e.g., linear feed-forward neural network, to predict emotion intensity, compound expressions, action units, facial expressions, and valence/arousal. Experimental results for five tasks from the sixth ABAW challenge demonstrate that our approach lets us significantly improve quality metrics on validation sets compared to existing non-ensemble *** Codes 68T10 Copyright © 2024, The Authors. All rights reserved.

关键词： Feedforward neural networks

来源：评论

学校读者我要写书评

暂无评论

Speaker-Aware Training of Speech Emotion Classifier with Speaker Recognition 23rd

Speaker-Aware Training of Speech Emotion Classifier with Spe...

引用

23rd International Conference on Speech and Computer, SPECOM 2021

作者： Savchenko, Lyudmila V. Savchenko, Andrey HSE University Laboratory of Algorithms and Technologies for Network Analysis Nizhny Novgorod Russia

ISBN: (纸本)9783030878016

In this paper, we discuss the possibility to improve the accuracy of speech emotion recognition in multi-user systems. We assume that a small corpus of speech emotional data is available for each speaker of interest. It is proposed to train a speaker-independent emotion classifier of arbitrary audio features including deep embeddings and fine-tune it on the utterances of each speaker. As a result, every user is associated with his or her own emotion recognition model. The concrete fine-tuned classifier may be chosen using a speaker recognition algorithm or even fixed if the identity of a user is known. It is experimentally shown that the proposed approach makes it possible to significantly improve the quality of conventional speaker-independent emotion classifier. © 2021, Springer Nature Switzerland AG.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：