
Research Square (Research Square), Journal Year: 2023, Volume and Issue: unknown
Published: Nov. 6, 2023
Language: Английский
Research Square (Research Square), Journal Year: 2023, Volume and Issue: unknown
Published: Nov. 6, 2023
Language: Английский
Multimedia Tools and Applications, Journal Year: 2023, Volume and Issue: 83(9), P. 26819 - 26870
Published: Sept. 5, 2023
Language: Английский
Citations
14Earth Science Informatics, Journal Year: 2025, Volume and Issue: 18(1)
Published: Jan. 1, 2025
Language: Английский
Citations
0Analog Integrated Circuits and Signal Processing, Journal Year: 2025, Volume and Issue: 123(2)
Published: March 24, 2025
Language: Английский
Citations
0Potato Research, Journal Year: 2025, Volume and Issue: unknown
Published: March 31, 2025
Language: Английский
Citations
0Acta Geophysica, Journal Year: 2024, Volume and Issue: 72(4), P. 2581 - 2596
Published: May 7, 2024
Language: Английский
Citations
2Published: Oct. 11, 2023
Video surveillance has been crucial for security in recent years, thus it's critical to guarantee the accuracy of these recordings. Unfortunately, it is easy fake films by erasing an object from a scene and leaving no visual evidence. In this study, approach video intra-frame forgery forensics based on SSIM (Structural Similarity Index Measure) proposed. This system can recognize forged frames automatically. order extract steganography features, method first decompresses input into sequence frames. It next calculates each frame's motion residual map. The ability tell whether or not deleted key problem security.
Language: Английский
Citations
6Published: April 26, 2024
The rampant spread of misinformation on online platforms has prompted extensive research into effective fake news detection methods. This paper conducts a comprehensive review various techniques utilized in this domain, with specific emphasis natural language processing (NLP), machine learning, and deep learning models. Central to these methods is the meticulous analysis source credibility contextual information. By scrutinizing origins stories considering surrounding context, strive differentiate between deceptive narratives genuine content. NLP play pivotal role, enabling extraction subtle linguistic patterns that aid identification misleading Moreover, underscores critical importance factual verification ensuring accurate detection. Factual involves cross-referencing information reliable sources, confirming authenticity claims made stories. integrating insights from analysis, understanding, capabilities advanced models, provide robust frameworks for tackling challenges posed by news.
Language: Английский
Citations
1Earth Science Informatics, Journal Year: 2024, Volume and Issue: 17(5), P. 4575 - 4595
Published: July 15, 2024
Language: Английский
Citations
0Journal of Advances in Information Technology, Journal Year: 2024, Volume and Issue: 8, P. 982 - 990
Published: Jan. 1, 2024
Language: Английский
Citations
0Artificial Intelligence Review, Journal Year: 2024, Volume and Issue: 58(2)
Published: Dec. 20, 2024
Satellite images often have very narrow brightness value ranges, so it is necessary to enhance the contrast and brightness, maintain quality of visual information, preserve pertinent details in before conducting additional analysis. This because improving crucial image processing analysis as makes easier for people identify comprehend images. The Incomplete Beta Function (IBF) a popular transformation function Image Contrast Enhancement (ICE). Nevertheless, IBF has modest efficiency parameter selection, small set adjustable parameters stretching regions with high or low gray levels, enhancement almost ineffective at either end. Meta-heuristic algorithms been utilized efficiently effectively over past few decades solve complicated problems. paper presents an Augmented version Elk Herd Optimizer (AEHO) combined other traditional ICE techniques improve edge details, entropy, local contrast, low-contrast natural satellite AEHO method employs multi-stage strategic procedure, where its mathematical model undergoes several enhancements being applied allow further exploration exploitation features. uses pre-established fitness criterion purpose optimizing rework well-known effective assessment technique objective standard this purpose. In proposed model, limited adaptive histogram equalization was first prior step ameliorate color intensity. Then, optimal IBF's were adaptively determined using AEHO. After that, bilateral gamma correction used without sacrificing quality. AEHO-based tested on scenes, certain images, publicly available addition five built based pre-existing meta-heuristics, performance compared against state-of-the-art algorithms. evaluation achieved utilizing variety full-reference, no-reference, norms. experimental findings illustrated that can successfully outperform employed same conventional methods included comparison. results ten showed presented performs better than all comparative corresponding criteria terms average peak signal-to-noise ratio, universal index, structural contrast-quality values discrete entropy results, which are more 32.30, 94.0%, 0.98.9%, 7.4, respectively. nutshell, be efficient tackle
Language: Английский
Citations
0