Audio-Video Data Processing
winter semester, 2020-2021
Student groups: Computer-Science in Romanian 3rd year, Computer-Science in English 3rd year
Professors:
Lect. Dr. Adrian Sterca [forest at cs dot ubbcluj dot ro]
Objectives:
At the end of this course, students should have fundamental knowledge in:
- digital audio and video signal representation and processing
- audio-video codec standards like ISO's MPEG and ITU-T H.26[1234]
- audio-video streaming over best-effort networks
- satellite audio-video broadcasting
- computer vision intro
They should also have expertise in designing and implementing an audio-video
processing system.
Pentru a urmari cursurile si a participa la laborator, studentii trebuie sa se adauge
la team-ul Procesarea datelor audio-video 2020
din Microsoft Teams.
Studentii care nu au minim media 5 pe laboratoare (un laborator nepredat se puncteaza cu nota 1) si nu au ales sa faca proiect,
vor trebui sa dea un examen quiz.
Examenul quiz o sa aiba loc in saptamana a 14-a a semestrului, adica JOI, 14 IANUARIE 2021, ora 14.00, pe moodle.cs.ubbcluj.ro.
Joi, 14 ianuarie 2021, la ora 14.00 o sa ne intalnim pe Ms Teams pentru a discuta detaliile despre examen si pentru a da examenul
efectiv. Examenul in sine o sa fie quiz de 30 de minute cu intrebari din slide-urile cursului (postate pe aceasta pagina).
Va trebui sa va inrolati in prealabil la cursul 'Procesarea datelor audio-video' pe moodle.cs.ubbcluj.ro.
The above announcement in English:
Students who do not obtain a minimum average grade of 5 for the labs (a lab not shown to the
professor will be graded with the grade 1) and did not choose the project evaluation criterion will have to take a quiz exam.
The quiz exam will take place during the last PDAV course of the semester (week 14 of the semester; 14th January 2021, 14.00 hours)
on moodle.cs.ubbcluj.ro. Thursday, 14th of January 2021, at 14.00 hours we will meet on Ms Teams to discuss the exam details
and to actually take the exam. The quiz exam will last 30 minutes and will contain questions from the course's slides
(posted on this web page). You should enroll in advance to the 'Procesarea datelor audio-video' course on moodle.cs.ubbcluj.ro.
Proiectele se vor preda de catre cei care au ales sa faca proiect (si au discutat cu mine despre proiect)
Marti, 19 Ianuarie 2021, ora 11.00 pe Ms Teams.
News [23-February-2021]: Note finale PDAV
Daca aveti neclaritati sau observati erori in notare, va rog sa imi scrieti mail.
Course description:
- Introduction
- Introduction to multimedia
- Color spaces
- Analog representation of audio and video signals
- Digital teprezentation of still images. The JPEG compression standard
- Digital representation of audio signal
- Audio and video formats (containers), codecs
- AV containers: .avi, .ogg, .mp4, .vob, .3gp, .mkv
- MPEG-1,2,4
- H.264/MPEG-4 AVC and the SVC extension
- H.265, High Efficiency Video Coding (HEVC)
- VP9
- Streaming and signalling protocols
- RTP/RTCP - Real Time Protocol
- RTSP - Real Time Streaming Protocol
- SDP - Session Description Protocol
- SIP - Session Initiation Protocol
- HTTP Streaming, DASH
- Congestion control algorithms for audio-video applications in "best-effort" networks
- TCP AIMD (Additive Increase Multiplicative Decrease)
- DCCP - Datagram Congestion Control Protocol
- TFRC - TCP-Friendly Rate Control
- UTFRC - Utility-driven TCP-Friendly Rate Control
- Voice over IP
- Audio-video communication in satellite networks
- Satellite communications
- Frequency bands
- DBS - Direct Broadcast Satellite (Direct-To-Home)
- DVB standards: DVB-S (satellite), DVB-T (terrestrial), DVB-C (cable)
- The FFMpeg library
- libavcodec
- libavformat
- libswscale
- libavfilter
- Audio-video for the web
- Audio-Video support in HTML 5
- Real-Time Communication on the web (WebRTC)
- Introduction to Computer Vision
- What is Computer Vision ?
- Feature extraction & feature matching
- Image features: SIFT
- Image features: SURF
- Multimedia QoS in Internet
- QoS terminology
- QoS requirements in "best-effort" networks
- P2P TV
- Peer-2-Peer networks
- DHT - Distributed Hash Tables
- Applications
Course Requirements (grading criteria):
There are 3 possible grading criteria and each student is allowed to choose the one that suits him best:
- Project: Students should develop a project related to
audio-video data processing either individually or in small teams of 2 persons. Students must choose
the project theme in the beginning of the semester and they must present the project at the end of the semester.
During the semester, they must also show work progress and intermediate versions of the project at the labs.
The possible grades a student can receive on the project are: 10, 9 and 4. No other grades are possible.
Examples of project ideas are here: List of possible projects (in romanian).
Please note that the theme of
each project and also its developement must be discussed with me at the laboratory because there are
several ways of implementing a specific task (like the ones mentioned in the list of projects from above), some
of them are very easy (using a library) and some of them are more complicated. Generally, I prefer that you
work on your project not relying on an additional external library, but rather do the low-level stuff
yourself.
- Labs: Students must complete 3 laboratory tasks (i.e. 3 labs) during the
semester. The lab tasks require the students to build a part of a video codec
and are detailed here: Laboratory tasks.
The final grade for this course is the average lab grade.
- Quiz test: If the student does not get an average lab grade of at least 5 and he/she
does not do a project, he/she must take the quiz test and the grade he/she receives for this quiz test is the final grade. This grade
can not be greater than 6.