Canonical Correlation Analysis Feature Fusion With Patch of Interest: A Dynamic Local Feature Matching for Face Sketch Image Retrieval

Abstract

An automatic photo retrieval system based on a face sketch has very useful application as to narrow down potential suspects in criminal investigations. This is true when there is no other evident except the face sketch that is rendered based on the recollection of a victim or eyewitness. Among the noticeable difficulties in matching the sketch and photo due to its modality difference are the generated sketch has some tendency of shape exaggeration, the sketch has very less accurate details and the real-world photo may expose to lighting variation unlike the sketch. In this paper, we attempt to address these complications by matching the sketch and photos using dynamic local feature of Difference of Gaussian Oriented Gradient Histogram (DoGOGH) on some selected patches. To avoid discriminative power degradation due to a large number of gallery images, two stage matching blocks are introduced in a cascaded fashion. The front block matches the feature such that it short lists k most similar photos for the second block. In this front block, Histogram of Oriented Gradient (HOG) and Gabor Wavelet (GW) features are fused by maximizing the correlation between the two using Canonical Correlation Analysis (CCA). Based on the short listed photos, the following block re-matched the sketch and photos using dynamically extracted local feature on its Patch of Interest (PoI). Eventually, the matching scores from the blocks are fused before getting rank-1 accuracy. The experimental results on two baseline datasets indicate that the proposed method outperforms the stateof-the-art methods. The extended evaluation on semi-forensic and forensic sketch datasets demonstrate its usage feasibility.

Performance evaluation of generative adversarial networks for generating mugshot images from text description

“…Face alignment places similar face components from distinct photos in the same location. In this study, the fiducial points that had been used during training and testing is similar to the study in [4]. Three fiducial points had been selected in this study, which are at the left and right face edge and chin tip.…”

“…From past studies, face sketch recognition could be categorized into two main categories which are intra-modality and inter-modality approaches. Intra-modality approaches attempt to synthesize sketch to the pseudo photo in order to recognize face sketches in the same modality (sketch or photo) and these approaches have been proven to have a great performance as compared to the inter-modality approaches [4]. Even though this existing method is effective in recognizing forensic sketches but it has one major drawback, which is the forensic sketch has to be created before the transformation task (i.e., synthesis) can be performed.…”

Application of Dynamic Image Fusion Technology in Development of Innovation Capabilities of College Students Majoring in Art Design

“…In order to achieve a better image fusion effect, it is necessary to denoise the low dynamic range image sequences acquired under different exposures before having them processed by the multi-exposure fusion algorithm. The commonly used denoising methods include bilateral filtering and non-local average filtering [29,30]. Figure 3 is the implementation framework for multi-exposure fusion.…”

Novel hybrid generative adversarial network for synthesizing image from sketch

“…Setumin et al [19] have addressed the problem of shape exaggeration of generalized sketch where a Gradient histogram on the basis of Gaussian value is designed for managing dynamic local features. A block-based operation is carried out for feature management using histogram of oriented gradient while the canonical correlation analysis is adopted to increase the correlation.…”