Av Recognition
Download as PPTX, PDF0 likes318 views
The document outlines an automatic video (AV) recognition system with the following key points: 1. The system architecture includes components like a client, server, and algorithms for processing, analyzing, and identifying AVs. 2. The recognition process involves inputting video, regularizing, filtering, comparing features, and summing up results. Techniques like FFT and second derivative analysis are used. 3. Examples of applying the system to song clips from movies are provided. The document concludes with releasing a version of the "AV-killer" system and posing questions about its design, evaluation, efficiency, and capabilities.
Av Recognition
- 1. AV recognition Teacher: We*-ch*** L**Presenter : P***A Hi**ng
- 3. AV!OUT!70%successful rate未來加強輪廓辨識，像是身形曲線、器官等，才能更精準提升過濾功能。Is this have more smart way to solve that?AV Killer讓Youtube笑了
- 4. ClientServerAlgorithmpyNumpyScipyMatplotPurpose scheme - System architectureTrainingProcessSODDB(MS SQL)MICIdentify ProcessAV’sslice
- 5. Purpose scheme – AlgorithmInputRegularSectionFilterFFT二次微分特徵比對Sum up
- 7. For example Mr. children – GIFTLinked Park – New divide Teriyaki boyz - Tokyo drift
- 8. For exampleSector 01Sector 02Sector 03Sector 04Sector 05
- 10. ConclusionTraining Video in database by “D:\”Finish the processes of “sector” and “filter”Releasing the AV-killer version α