• Home
  • Search
  • Can using a pre-trained deep learning model as the feature extractor in the bag-of-deep-visual-words model always improve image classification accuracy?
  • Cite Icon4
  • https://doi.org/10.1371/journal.pone.0298228Copy DOI Icon

Can using a pre-trained deep learning model as the feature extractor in the bag-of-deep-visual-words model always improve image classification accuracy?

Show More
  • Abstract
  • Highlights & Summary
  • PDF
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

This article investigates whether higher classification accuracy can always be achieved by utilizing a pre-trained deep learning model as the feature extractor in the Bag-of-Deep-Visual-Words (BoDVW) classification model, as opposed to directly using the new classification layer of the pre-trained model for classification. Considering the multiple factors related to the feature extractor -such as model architecture, fine-tuning strategy, number of training samples, feature extraction method, and feature encoding method—we investigate these factors through experiments and then provide detailed answers to the question. In our experiments, we use five feature encoding methods: hard-voting, soft-voting, locally constrained linear coding, super vector coding, and fisher vector (FV). We also employ two popular feature extraction methods: one (denoted as Ext-DFs(CP)) uses a convolutional or non-global pooling layer, and another (denoted as Ext-DFs(FC)) uses a fully-connected or global pooling layer. Three pre-trained models—VGGNet-16, ResNext-50(32×4d), and Swin-B—are utilized as feature extractors. Experimental results on six datasets (15-Scenes, TF-Flowers, MIT Indoor-67, COVID-19 CXR, NWPU-RESISC45, and Caltech-101) reveal that compared to using the pre-trained model with only the new classification layer re-trained for classification, employing it as the feature extractor in the BoDVW model improves the accuracy in 35 out of 36 experiments when using FV. With Ext-DFs(CP), the accuracy increases by 0.13% to 8.43% (averaged at 3.11%), and with Ext-DFs(FC), it increases by 1.06% to 14.63% (averaged at 5.66%). Furthermore, when all layers of the pre-trained model are fine-tuned and used as the feature extractor, the results vary depending on the methods used. If FV and Ext-DFs(FC) are used, the accuracy increases by 0.21% to 5.65% (averaged at 1.58%) in 14 out of 18 experiments. Our results suggest that while using a pre-trained deep learning model as the feature extractor does not always improve classification accuracy, it holds great potential as an accuracy improvement technique.

Loading PDF

Similar Papers
  • Conference Article
  • Citations1

Person re-identification with pre-trained deep learning models and attribute correlations

  • Nov 01, 2016
  • Ngoc-Bao Nguyen +3
  • Abstract
  • Citations2

Comparison of pre-trained deep learning model classification performance of COVID-19 and normal chest X-ray images

  • Dec 01, 2022
  • Journal of Medical Imaging and Radiation Sciences
  • Yudthaphon Vichianin +8
  • Research Article
  • Citations115

Exploiting pretrained CNN models for the development of an EEG-based robust BCI framework

  • Jan 25, 2022
  • Computers in Biology and Medicine
  • Muhammad Tariq Sadiq +5
  • PDF
  • Research Article
  • Citations53

Enhancing coffee bean classification: a comparative analysis of pre-trained deep learning models

  • Apr 01, 2024
  • Neural Computing and Applications
  • Esraa Hassan
  • Research Article
  • Citations3

Open-source skull reconstruction with MONAI

  • Jun 14, 2023
  • SoftwareX
  • Jianning Li +9
  • Research Article

Multi-Model Feature Extraction and Classification of Steel Rods Using Fuzzy Logic Systems

  • Mar 29, 2025
  • Journal of Information Systems Engineering and Management
  • Chandrashekar P, G.N.K Suresh Babu
  • Research Article

An automated tuberculosis detection approach using deep learning and machine learning techniques from chest X-ray images: a step towards effective diagnosis

  • Feb 02, 2026
  • Polish Journal of Radiology
  • Abu Saleh Mohammad Nabil +7
  • Conference Article
  • Citations24

Chest X-Ray Image to Classify Lung diseases in Different Resolution Size using DenseNet-121 Architectures

  • Sep 13, 2021
  • Ovy Rochmawanti +1
  • Research Article

Integrating Few-Shot Learning and Pre-trained Models into Similarity-Based PHM using Small Data in Complex Engineering Systems

  • Oct 26, 2025
  • Annual Conference of the PHM Society
  • Takanobu Minami +1
  • Research Article
  • Citations3

EXPLORING TRANSFER LEARNING IN IMAGE ANALYSIS USING FEATURE EXTRACTION WITH PRE-TRAINED MODELS

  • Feb 01, 2024
  • ICTACT Journal on Image and Video Processing
  • Kavita Wagh +3
  • PDF
  • Research Article
  • Citations10

Investigation of self-supervised pre-trained models for classification of voice quality from speech and neck surface accelerometer signals

  • Jul 28, 2023
  • Computer Speech & Language
  • Sudarsana Reddy Kadiri +2
  • Research Article
  • Citations73

Automated detection of pneumonia cases using deep transfer learning with paediatric chest X-ray images.

  • Apr 16, 2021
  • The British Journal of Radiology
  • Mohammad Salehi +4
  • Research Article
  • Citations44

Image cyberbullying detection and recognition using transfer deep machine learning

  • Dec 13, 2023
  • International Journal of Cognitive Computing in Engineering
  • Ammar Almomani +5
  • Conference Article
  • Citations6

Transfer Learning for Maritime Vessel Detection using Deep Neural Networks

  • Sep 19, 2021
  • Fahimeh Farahnakian +2
  • Research Article
  • Citations1

Enhancing Object Detection Techniques Through Transfer Learning and Pre-trained Models

  • Sep 30, 2024
  • Journal of engineering sciences and information technology
  • Ashwaq Katham Mtashre +2
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.