Matches in SemOpenAlex for { <https://semopenalex.org/work/W4226054261> ?p ?o ?g. }
Showing items 1 to 90 of
90
with 100 items per page.
- W4226054261 endingPage "238" @default.
- W4226054261 startingPage "227" @default.
- W4226054261 abstract "AbstractIn the age of technologies, a brand-new technology in the field of deep neural networks and artificial intelligence has been developed. This technology of creating counter-fit realistic videos is known as Deepfake and is born lately. It's troublesome to discriminate between fake and real videos which are daily uploaded on the web across different websites. There are several Deepfake open-source creation methods available online, which results in a growing number of fake videos on the web. There are many quick and efficient ways and techniques which are made to spot, observe, and detect these developments. To define a few, comparison of backgrounds, facial artifacts, blinking of eyes, pattern analysis, pose, and likewise features of the face and surroundings are used to help in detecting a Deepfake video. In this paper, a simple but effective approach for detecting fakes using 2D Convolutional Neural Network (Conv2D) is followed and the use of 3D Convolutional Neural Network (Conv3D) is introduced in Deepfake Detection with the addition of Multi-Task Cascade Convolutional Neural Network (MTCNN) in preprocessing. The significance and importance of this work are that a different approach is taken in preprocessing which in return gives better results and the introduction for the use of Conv3D. The proposed models are trained and tested on celeb-df v2 and Deepfake-TIMIT both low-quality (LQ) and high-quality videos (HQ) with gaining the highest accuracy of 99.81%. Also, compared to the state-of-the-art techniques, the proposed approach helped in gaining a good accuracy on low-quality videos, and the difference between both HQ and LQ is observed to be less than 1%.KeywordsDeepfake detection,3D CNN,Video forgery analysisDigital forensics" @default.
- W4226054261 created "2022-05-05" @default.
- W4226054261 creator A5072106317 @default.
- W4226054261 creator A5089230620 @default.
- W4226054261 date "2022-01-01" @default.
- W4226054261 modified "2023-09-27" @default.
- W4226054261 title "A Simple and Effective Way to Detect DeepFakes: Using 2D and 3D CNN" @default.
- W4226054261 cites W1482998036 @default.
- W4226054261 cites W2891145043 @default.
- W4226054261 cites W2911424785 @default.
- W4226054261 cites W2913399670 @default.
- W4226054261 cites W2914447220 @default.
- W4226054261 cites W2962958939 @default.
- W4226054261 cites W2963684180 @default.
- W4226054261 cites W2963720850 @default.
- W4226054261 cites W2995516027 @default.
- W4226054261 cites W3011236737 @default.
- W4226054261 cites W3012472557 @default.
- W4226054261 cites W3015341856 @default.
- W4226054261 cites W3017837134 @default.
- W4226054261 cites W3023887666 @default.
- W4226054261 cites W3034713808 @default.
- W4226054261 cites W3036381647 @default.
- W4226054261 cites W3036626178 @default.
- W4226054261 cites W3043392332 @default.
- W4226054261 cites W3083246145 @default.
- W4226054261 cites W3086623482 @default.
- W4226054261 cites W3094477634 @default.
- W4226054261 cites W3128565165 @default.
- W4226054261 cites W3183175034 @default.
- W4226054261 doi "https://doi.org/10.1007/978-981-16-9089-1_19" @default.
- W4226054261 hasPublicationYear "2022" @default.
- W4226054261 type Work @default.
- W4226054261 citedByCount "2" @default.
- W4226054261 countsByYear W42260542612022 @default.
- W4226054261 crossrefType "book-chapter" @default.
- W4226054261 hasAuthorship W4226054261A5072106317 @default.
- W4226054261 hasAuthorship W4226054261A5089230620 @default.
- W4226054261 hasConcept C111472728 @default.
- W4226054261 hasConcept C119857082 @default.
- W4226054261 hasConcept C136764020 @default.
- W4226054261 hasConcept C138885662 @default.
- W4226054261 hasConcept C144024400 @default.
- W4226054261 hasConcept C153180895 @default.
- W4226054261 hasConcept C154945302 @default.
- W4226054261 hasConcept C202444582 @default.
- W4226054261 hasConcept C2779304628 @default.
- W4226054261 hasConcept C2779530757 @default.
- W4226054261 hasConcept C33923547 @default.
- W4226054261 hasConcept C34736171 @default.
- W4226054261 hasConcept C36289849 @default.
- W4226054261 hasConcept C41008148 @default.
- W4226054261 hasConcept C71901391 @default.
- W4226054261 hasConcept C81363708 @default.
- W4226054261 hasConcept C9652623 @default.
- W4226054261 hasConceptScore W4226054261C111472728 @default.
- W4226054261 hasConceptScore W4226054261C119857082 @default.
- W4226054261 hasConceptScore W4226054261C136764020 @default.
- W4226054261 hasConceptScore W4226054261C138885662 @default.
- W4226054261 hasConceptScore W4226054261C144024400 @default.
- W4226054261 hasConceptScore W4226054261C153180895 @default.
- W4226054261 hasConceptScore W4226054261C154945302 @default.
- W4226054261 hasConceptScore W4226054261C202444582 @default.
- W4226054261 hasConceptScore W4226054261C2779304628 @default.
- W4226054261 hasConceptScore W4226054261C2779530757 @default.
- W4226054261 hasConceptScore W4226054261C33923547 @default.
- W4226054261 hasConceptScore W4226054261C34736171 @default.
- W4226054261 hasConceptScore W4226054261C36289849 @default.
- W4226054261 hasConceptScore W4226054261C41008148 @default.
- W4226054261 hasConceptScore W4226054261C71901391 @default.
- W4226054261 hasConceptScore W4226054261C81363708 @default.
- W4226054261 hasConceptScore W4226054261C9652623 @default.
- W4226054261 hasLocation W42260542611 @default.
- W4226054261 hasOpenAccess W4226054261 @default.
- W4226054261 hasPrimaryLocation W42260542611 @default.
- W4226054261 hasRelatedWork W2175746458 @default.
- W4226054261 hasRelatedWork W2521062615 @default.
- W4226054261 hasRelatedWork W2572787276 @default.
- W4226054261 hasRelatedWork W2976316911 @default.
- W4226054261 hasRelatedWork W2977314777 @default.
- W4226054261 hasRelatedWork W3027997911 @default.
- W4226054261 hasRelatedWork W3081496756 @default.
- W4226054261 hasRelatedWork W3093612317 @default.
- W4226054261 hasRelatedWork W3124066309 @default.
- W4226054261 hasRelatedWork W4287776258 @default.
- W4226054261 isParatext "false" @default.
- W4226054261 isRetracted "false" @default.
- W4226054261 workType "book-chapter" @default.