FAUST: Fast Per-Scene Encoding Using Entropy-Based Scene Detection and Machine Learning
HTTP adaptive video streaming is a widespread and sought-after technology on the Internet that allows clients to dynamically switch between different stream qualities presented in the bitrate ladder to optimize overall received video quality. Currently, there exist several approaches of different co...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
FRUCT
2021-10-01
|
| Series: | Proceedings of the XXth Conference of Open Innovations Association FRUCT |
| Subjects: | |
| Online Access: | https://www.fruct.org/publications/fruct30/files/Zab.pdf |
| _version_ | 1819198460269690880 |
|---|---|
| author | Anatoliy Zabrovskiy Prateek Agrawal Christian Timmerer Radu Prodan |
| author_facet | Anatoliy Zabrovskiy Prateek Agrawal Christian Timmerer Radu Prodan |
| author_sort | Anatoliy Zabrovskiy |
| collection | DOAJ |
| description | HTTP adaptive video streaming is a widespread and sought-after technology on the Internet that allows clients to dynamically switch between different stream qualities presented in the bitrate ladder to optimize overall received video quality. Currently, there exist several approaches of different complexity for building such a ladder. The simplest method is to use a static bitrate ladder, and the more complex one is to compute a per-title encoding ladder. The main drawback of these approaches is that they do not provide bitrate ladders for scenes with different visual complexity within the video. Moreover, most modern methods require additional computationally-intensive test encodings of the entire video to construct the convex hull, used to calculate the bitrate ladder. This paper proposes a new fast per-scene encoding approach called FAUST based on 1) quick entropy-based scene detection and 2) prediction of optimized bitrate ladder for each scene using an artificial neural network. The results show that our model reduces the mean absolute error to 0.15, the mean square error to 0.08, and the bitrate to 13.5% while increasing the difference in video multimethod assessment fusion to 5.6 points. |
| first_indexed | 2024-12-23T03:00:47Z |
| format | Article |
| id | doaj.art-2c4ff9718dd84364864e573ee8909ef5 |
| institution | Directory Open Access Journal |
| issn | 2305-7254 2343-0737 |
| language | English |
| last_indexed | 2024-12-23T03:00:47Z |
| publishDate | 2021-10-01 |
| publisher | FRUCT |
| record_format | Article |
| series | Proceedings of the XXth Conference of Open Innovations Association FRUCT |
| spelling | doaj.art-2c4ff9718dd84364864e573ee8909ef52022-12-21T18:02:26ZengFRUCTProceedings of the XXth Conference of Open Innovations Association FRUCT2305-72542343-07372021-10-0130129230410.23919/FRUCT53335.2021.9599963FAUST: Fast Per-Scene Encoding Using Entropy-Based Scene Detection and Machine LearningAnatoliy Zabrovskiy0Prateek Agrawal1Christian Timmerer2Radu Prodan3University of Klagenfurt, AustriaLovely Professional University, IndiaLovely Professional University, IndiaLovely Professional University, IndiaHTTP adaptive video streaming is a widespread and sought-after technology on the Internet that allows clients to dynamically switch between different stream qualities presented in the bitrate ladder to optimize overall received video quality. Currently, there exist several approaches of different complexity for building such a ladder. The simplest method is to use a static bitrate ladder, and the more complex one is to compute a per-title encoding ladder. The main drawback of these approaches is that they do not provide bitrate ladders for scenes with different visual complexity within the video. Moreover, most modern methods require additional computationally-intensive test encodings of the entire video to construct the convex hull, used to calculate the bitrate ladder. This paper proposes a new fast per-scene encoding approach called FAUST based on 1) quick entropy-based scene detection and 2) prediction of optimized bitrate ladder for each scene using an artificial neural network. The results show that our model reduces the mean absolute error to 0.15, the mean square error to 0.08, and the bitrate to 13.5% while increasing the difference in video multimethod assessment fusion to 5.6 points.https://www.fruct.org/publications/fruct30/files/Zab.pdfper-title encodingper-shot encodingscene detectionvideo entropyneural networksvideo encodingvideo transcodinghttp adaptive streamingmpeg-dash |
| spellingShingle | Anatoliy Zabrovskiy Prateek Agrawal Christian Timmerer Radu Prodan FAUST: Fast Per-Scene Encoding Using Entropy-Based Scene Detection and Machine Learning Proceedings of the XXth Conference of Open Innovations Association FRUCT per-title encoding per-shot encoding scene detection video entropy neural networks video encoding video transcoding http adaptive streaming mpeg-dash |
| title | FAUST: Fast Per-Scene Encoding Using Entropy-Based Scene Detection and Machine Learning |
| title_full | FAUST: Fast Per-Scene Encoding Using Entropy-Based Scene Detection and Machine Learning |
| title_fullStr | FAUST: Fast Per-Scene Encoding Using Entropy-Based Scene Detection and Machine Learning |
| title_full_unstemmed | FAUST: Fast Per-Scene Encoding Using Entropy-Based Scene Detection and Machine Learning |
| title_short | FAUST: Fast Per-Scene Encoding Using Entropy-Based Scene Detection and Machine Learning |
| title_sort | faust fast per scene encoding using entropy based scene detection and machine learning |
| topic | per-title encoding per-shot encoding scene detection video entropy neural networks video encoding video transcoding http adaptive streaming mpeg-dash |
| url | https://www.fruct.org/publications/fruct30/files/Zab.pdf |
| work_keys_str_mv | AT anatoliyzabrovskiy faustfastpersceneencodingusingentropybasedscenedetectionandmachinelearning AT prateekagrawal faustfastpersceneencodingusingentropybasedscenedetectionandmachinelearning AT christiantimmerer faustfastpersceneencodingusingentropybasedscenedetectionandmachinelearning AT raduprodan faustfastpersceneencodingusingentropybasedscenedetectionandmachinelearning |