DeepGUM: Learning Deep Robust Regression with a Gaussian-Uniform Mixture Model

被引:8
|
作者
Lathuiliere, Stephane [1 ,3 ]
Mesejo, Pablo [1 ,2 ]
Alameda-Pineda, Xavier [1 ]
Horaud, Radu [1 ]
机构
[1] Inria Grenoble Rhone Alpes, Montbonnot St Martin, France
[2] Univ Granada, Granada, Spain
[3] Univ Trento, Trento, Italy
来源
COMPUTER VISION - ECCV 2018, PT V | 2018年 / 11209卷
基金
欧洲研究理事会;
关键词
Robust regression; Deep neural networks; Mixture model; Outlier detection; HEAD-POSE ESTIMATION;
D O I
10.1007/978-3-030-01228-1_13
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In this paper we address the problem of how to robustly train a ConvNet for regression, or deep robust regression. Traditionally, deep regression employ the L-2 loss function, known to be sensitive to outliers, i.e. samples that either lie at an abnormal distance away from the majority of the training samples, or that correspond to wrongly annotated targets. This means that, during back-propagation, outliers may bias the training process due to the high magnitude of their gradient. In this paper, we propose DeepGUM: a deep regression model that is robust to outliers thanks to the use of a Gaussian-uniform mixture model. We derive an optimization algorithm that alternates between the unsupervised detection of outliers using expectation-maximization, and the supervised training with cleaned samples using stochastic gradient descent. DeepGUM is able to adapt to a continuously evolving outlier distribution, avoiding to manually impose any threshold on the proportion of outliers in the training set. Extensive experimental evaluations on four different tasks (facial and fashion landmark detection, age and head pose estimation) lead us to conclude that our novel robust technique provides reliability in the presence of various types of noise and protection against a high percentage of outliers.
引用
收藏
页码:205 / 221
页数:17
相关论文
共 50 条
  • [41] Joint Federated Learning Using Deep Segmentation and the Gaussian Mixture Model for Breast Cancer Tumors
    Tan, Y. Nguyen
    Lam, Pham Duc
    Tinh, Vo Phuc
    Le, Duy-Dong
    Nam, Nguyen Hoang
    Khoa, Tran Anh
    IEEE ACCESS, 2024, 12 : 94231 - 94249
  • [42] Deep learning convolutional neural network (CNN) With Gaussian mixture model for predicting pancreatic cancer
    Kaushik Sekaran
    P. Chandana
    N. Murali Krishna
    Seifedine Kadry
    Multimedia Tools and Applications, 2020, 79 : 10233 - 10247
  • [43] Deep Gaussian mixture model based instance relevance estimation for multiple instance learning applications
    Waqas, Muhammad
    Tahir, Muhammad Atif
    Qureshi, Rizwan
    APPLIED INTELLIGENCE, 2023, 53 (09) : 10310 - 10325
  • [44] Moving Vehicle Detection and Classification Using Gaussian Mixture Model and Ensemble Deep Learning Technique
    Jagannathan, Preetha
    Rajkumar, Sujatha
    Frnda, Jaroslav
    Divakarachari, Parameshachari Bidare
    Subramani, Prabu
    WIRELESS COMMUNICATIONS & MOBILE COMPUTING, 2021, 2021
  • [45] Deep Gaussian mixture model based instance relevance estimation for multiple instance learning applications
    Muhammad Waqas
    Muhammad Atif Tahir
    Rizwan Qureshi
    Applied Intelligence, 2023, 53 : 10310 - 10325
  • [46] Integrating Regression model with Gaussian Mixture model for Image Super-Resolution
    Deepak, A. V. S.
    Ghanekar, Umesh
    2017 INTERNATIONAL CONFERENCE ON INTELLIGENT COMPUTING AND CONTROL SYSTEMS (ICICCS), 2017, : 1281 - 1286
  • [47] Matrix Variate Gaussian Mixture Distribution Steered Robust Metric Learning
    Luo, Lei
    Huang, Heng
    THIRTY-SECOND AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTIETH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE / EIGHTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2018, : 3722 - 3729
  • [48] Mixed Deep Gaussian Mixture Model: a clustering model for mixed datasets
    Fuchs, Robin
    Pommeret, Denys
    Viroli, Cinzia
    ADVANCES IN DATA ANALYSIS AND CLASSIFICATION, 2022, 16 (01) : 31 - 53
  • [49] Joint Gaussian mixture model for versatile deep visual model explanation
    Xie, Zhouyang
    He, Tianxiang
    Tian, Shengzhao
    Fu, Yan
    Zhou, Junlin
    Chen, Duanbing
    KNOWLEDGE-BASED SYSTEMS, 2023, 280
  • [50] Mixed Deep Gaussian Mixture Model: a clustering model for mixed datasets
    Robin Fuchs
    Denys Pommeret
    Cinzia Viroli
    Advances in Data Analysis and Classification, 2022, 16 : 31 - 53