Experimental evaluation of unsupervised image retrieval application using hybrid feature extraction by integrating deep learning and handcrafted techniques

Sudheer, Devulapalli; Potti, Anupama; Krishnan, Rajakumar; Khan, Md. Sameeruddin

doi:10.1016/j.matpr.2021.04.326

Cited by 13 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The matching of feature resemblance was achieved by picking a similarity measure using Approximate Nearest Neighbor (Annoy) Indexing, with time and precision adjusted. [25] proposed built a feature vector for Image Retrieval by combining deep learning and handcraft features. They act as a pre-trained feature extractors and fewer feature dimensions than other deep learning models, so Googlenet was used.…”

Section: Deep Learningmentioning

confidence: 99%

An Overview of Content-Based Image Retrieval Methods and Techniques

Hadid,

Hussein,

Al-Qaysi

et al. 2023

ijcsm

View full text Add to dashboard Cite

With the development of Internet technology and the popularity of digital devices, Content-Based Image Retrieval (CBIR) has been quickly developed and applied in various fields related to computer vision and artificial intelligence. Currently, it is possible to retrieve related images effectively and efficiently from a large-scale database with an input image. In the past ten years, great efforts have been made for new theories and models of CBIR, and many effective CBIR algorithms have been established. Content-based image retrieval helps to discover identical images in a big dataset that match a query image. The query image's representative feature similarities to the dataset images typically assist in ranking the images for retrieval. There are various past studies on different handicraft feature descriptors according to the visual features that describe the images: color, texture, and shape. However, deep learning has been the dominant alternative to manually planned feature engineering; it automatically takes the features from the data. The current work reviews recent advancements in content-based image retrieval. For a deeper understanding of the advancement, the explanation of current state-of-the-art approaches from various vantage points is also conducted. This review employs a taxonomy encompassing various retrieval networks, classification types, and descriptors and this study will help researchers make more progress in image retrieval

show abstract

Section: Deep Learningmentioning

confidence: 99%

An Overview of Content-Based Image Retrieval Methods and Techniques

Hadid,

Hussein,

Al-Qaysi

et al. 2023

ijcsm

View full text Add to dashboard Cite

show abstract

“…The combination of handcrafted features with CNN's features is becoming a popular approach to address various problems in different domains such as image scene geometry recognition [12], classification of working condition in froth flotation [13], signal gesture recognition [14], land cover content-based image retrieval [15], and pedestrian detection [16]. Moreover, it has been widely used in the medical field to perform classification [17], segmentation [12], and detection [18] tasks.…”

Section: Introductionmentioning

confidence: 99%

Deep learning with multiresolution handcrafted features for brain MRI segmentation

Mecheter

Abbod

Amira

et al. 2022

Artificial Intelligence in Medicine

View full text Add to dashboard Cite

“…However, when the training set is small, over‐fitting is prone to occur. [ 23,24 ] ResNet uses the identity mapping in the residual network to solve the problem of gradient disappearance or explosion in the feed‐forward/feedback propagation algorithm, but the training time is long. [ 25,26 ] Since CNN can extract effective features from data, and LSTM can discover the dependence relationship between data, we also discuss CNN‐LSTM model, but the combination of the two structures increases the operation cost.…”

Section: Introductionmentioning

confidence: 99%

A new method for Raman spectral analysis: Decision fusion‐based transfer learning model

Chen

Zhu

et al. 2022

J Raman Spectroscopy

View full text Add to dashboard Cite

As an emerging technology for artificial intelligence‐aided medical diagnosis, deep learning combined with Raman spectroscopy has great potential. The technology still has some problems in the actual medical diagnosis research process. The differences in spectrometers, experimental conditions, and experimental operations can result in non‐uniform and universally applicable data standards, which in turn lead to low data utilization. At the same time, it is still necessary to retrain the models when building diagnostic models for different diseases, which is time‐consuming and laborious. In this paper, a more complete transfer learning model for multiple types of serum Raman spectra is established for the first time, and a decision fusion strategy is applied to this diagnostic model. The Raman spectral data of serum from hepatitis B patients/control group, serum from abnormal thyroid function patients/control group, and serum from glioma patients/control group were selected as the source domains, and the Raman spectral data of tissue from hepatitis C patients/control group, serum from esophageal cancer patients/control group, and tissue from cervical cancer and cervical inflammation (patients/control) group were selected as the target domains. Three deep neural network models, ResNet, GoogLeNet, and CNN‐LSTM were trained in the source domain data for disease diagnosis, and the trained models were transfer to the target domain. The model is fine‐tuned by freezing different layers and then combined with logistic regression algorithms to construct a decision fusion model, which further improves the model effect. The results show that the proposed method can effectively improve the accuracy of transfer learning models. At the same time, this experiment extends the application of transfer learning in Raman spectroscopy and demonstrates that unrelated and scale‐different Raman datasets are still intrinsically connected, which also lays the foundation for us to build more stable and data‐inclusive spectral transfer learning fusion models in the future.

show abstract

Experimental evaluation of unsupervised image retrieval application using hybrid feature extraction by integrating deep learning and handcrafted techniques

Cited by 13 publications

References 31 publications

An Overview of Content-Based Image Retrieval Methods and Techniques

An Overview of Content-Based Image Retrieval Methods and Techniques

Deep learning with multiresolution handcrafted features for brain MRI segmentation

A new method for Raman spectral analysis: Decision fusion‐based transfer learning model

Contact Info

Product

Resources

About