Deep CNN Combined With Relevance Feedback for Trademark Image Retrieval

Pinjarkar, Latika; Шарма, Маниша; Selot, Smita

doi:10.1515/jisys-2018-0083

Cited by 13 publications

(8 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such as by computing a new query vector [36], or by modifying the similarity measure in such a way that relevant images have a high similarity value [33], or trying to separate relevant and non-relevant images using pattern classification techniques such as Support Vector Machines [20], Decision Trees [24], Clustering [9], Random Forests [6] or Convolutional Neural Networks [48]. Our approach processes the user's feedback and separates relevant from non-relevant images by exploiting a CNN pre-trained on a large data set as in [31,49]. Here, we make an extensive investigation on different approaches to exploit the RF when a modified CNN is used both for feature extraction and for classifying images as relevant or non-relevant to the given query.…”

Section: Relevance Feedbackmentioning

confidence: 99%

“…A different group of approaches use a formulation of RF in terms of a pattern classification task, by using the relevant and non-relevant image sets to train popular learning algorithms such as SVMs [20], neural networks and self-organising maps [27,32]. Even CNNs can be re-purposed for this task as proposed in [31,[47][48][49] to move the non-relevant image away from the query image through a modification of the feature space using the backpropagation algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Convolutional neural networks for relevance feedback in content based image retrieval

Putzu

Piras

Giacinto

2020

Multimed Tools Appl

View full text Add to dashboard Cite

Given the great success of Convolutional Neural Network (CNN) for image representation and classification tasks, we argue that Content-Based Image Retrieval (CBIR) systems could also leverage on CNN capabilities, mainly when Relevance Feedback (RF) mechanisms are employed. On the one hand, to improve the performances of CBIRs, that are strictly related to the effectiveness of the descriptors used to represent an image, as they aim at providing the user with images similar to an initial query image. On the other hand, to reduce the semantic gap between the similarity perceived by the user and the similarity computed by the machine, by exploiting an RF mechanism where the user labels the returned images as being relevant or not concerning her interests. Consequently, in this work, we propose a CBIR system based on transfer learning from a CNN trained on a vast image database, thus exploiting the generic image representation that it has already learned. Then, the pre-trained CNN is also fine-tuned exploiting the RF supplied by the user to reduce the semantic gap. In particular, after the user’s feedback, we propose to tune and then re-train the CNN according to the labelled set of relevant and non-relevant images. Then, we suggest different strategies to exploit the updated CNN for returning a novel set of images that are expected to be relevant to the user’s needs. Experimental results on different data sets show the effectiveness of the proposed mechanisms in improving the representation power of the CNN with respect to the user concept of image similarity. Moreover, the pros and cons of the different approaches can be clearly pointed out, thus providing clear guidelines for the implementation in production environments.

show abstract

Section: Relevance Feedbackmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Convolutional neural networks for relevance feedback in content based image retrieval

Putzu

Piras

Giacinto

2020

Multimed Tools Appl

View full text Add to dashboard Cite

show abstract

“…After i had been increased by 1, the selected target was added to the set of coarse screening targets. Then, the PD was calculated by formula (2). If PD is smaller than TPM, the image similarity is high, and the i+1-th image should be added to the set of coarse screening targets; if PD is greater than the threshold, the image similarity is low, and the i+1-th image should not be added.…”

Section: Coarse Screeningmentioning

confidence: 99%

“…The thriving of information technology (IT) has elevated the demand and expectation of tourists for high-quality tourism services. To make the itinerary more convenient and personalized, tourists need to query and retrieve information about the tourist attractions of interest [1,2].…”

Section: Introductionmentioning

confidence: 99%

Retrieval of Multiple Spatiotemporally Correlated Images on Tourist Attractions Based on Image Processing

Lü¹,

Zhang²,

Liu³

et al. 2020

View full text Add to dashboard Cite

The thriving of information technology (IT) has elevated the demand for intelligent query and retrieval of information about the tourist attractions of interest, which are the bases for preparing convenient and personalized itineraries. To realize accurate and rapid query of tourist attraction information (not limited to text information), this paper proposes a spatiotemporal feature extraction method and a ranking and retrieval method for multiple spatiotemporally correlated images (MSCIs) on tourist attractions based on deeply recursive convolutional network (DRCN). Firstly, the authors introduced the acquisition process of candidate spatiotemporally correlated images on tourist attractions, including both coarse screening and fine screening. Next, the workflow of spatiotemporal feature extraction from tourist attraction images was explained, as well as he proposed convolutional long short-term memory (ConvLSTM) algorithm. After that, the ranking model of MSCIs was constructed and derived. Experimental results demonstrate that our strategy is effective in the retrieval of tourist attraction images. The research results shed light on the fast and accurate retrieval of other types of images.

show abstract

“…A huge amount of trademark images are transmitted via the internet. In this situation, there is a growing interest in trademark image retrieval [1]- [10] and trademark shape recognition [11]- [17]. How to represent two (or more) separated shapes sufficiently in a trademark image is an important issue [18].…”

Section: Introductionmentioning

confidence: 99%

A Novel Shape Representation Method for Complex Trademark Image

Cai

Lan

et al. 2019

IEEE Access

View full text Add to dashboard Cite

Trademark orthogonal representation is an important research field in artificial intelligence technology and has extensive applications in trademark retrieval and recognition. However, the Gibbs phenomenon arises when complex trademark shapes are represented by using traditional methods, such as continuous orthogonal function and wavelet. In order to represent the complex trademark without Gibbs phenomenon, a novel method named hierarchical V-system (HV system) is proposed in this study, which is generated by V-system of multi-degree k (k = 0, 1, 2, 3). The hierarchical structure brings more detailed shape representation information. We show that the proposed method in this paper can represent the trademark shapes with fewer finite terms and obtain accurate representation result. This study also proposes the normalized descriptors of hierarchical V-system (HV descriptors). Furthermore, we demonstrate that the HV descriptors satisfy the invariance in rotation, translation, and scale transform. The experimental results show that the hierarchical V-system method can give reasonable descriptors for representing complex trademark without Gibbs phenomenon.

show abstract

Deep CNN Combined With Relevance Feedback for Trademark Image Retrieval

Cited by 13 publications

References 21 publications

Convolutional neural networks for relevance feedback in content based image retrieval

Convolutional neural networks for relevance feedback in content based image retrieval

Retrieval of Multiple Spatiotemporally Correlated Images on Tourist Attractions Based on Image Processing

A Novel Shape Representation Method for Complex Trademark Image

Contact Info

Product

Resources

About