Introduction
This page shows the results from the student projects that were part of GSoC 2020.
Projects
ABR module for FFmpeg
Description:
Add an Adaptive Bit Rate (ABR) module to FFmpeg.
More info from the student here: https://spartazhc.github.io/2020/08/23/ABR-meets-FFmpeg-in-detail/
Results: Successful
The throughput-based ABR algorithm was implemented
Student: Hongcheng Zhong
Mentor: Jun Zhao, Steven Liu
Audio Tones Source Filter
Description:
The aim of this project was to create an audio tones source filter that generates music algorithmically. The resulting filter uses Fluidsynth API and gives the user command over the instrumentation, pitch mapping, time control and style of the algorithmic music.
Results: Successful
Student: Ashutosh Pradhan
Mentor: Paul B Mahol
Backup Mentor: Thilo Borgmann
FLIF Encoder & Decoder Project
Description:
This was actually an umbrella, 2-part project consisting two students' work
Free Lossless Image Format, or FLIF is the precursor to JPEG XL and FUIF. It claims to have a higher level of compression or lower file size than conventional lossless image formats like PNG, Lossless JPEG 2000, etc. FLIF and FUIF development has been stopped in favour of JPEG XL.
The task was to write an encoder and decoder for this format for FFmpeg.
Results: Successful
Student: Anamitra Ghorui (And Kartik K. Khullar)
Mentor: darkapex
Backup Mentor: Thilo Borgmann
Implementation of a FLIF Codec in libavcodec
Description:
This was actually an umbrella, 2-part project consisting two students' work
FLIF is a modern lossless image format with very good compression ratio compared to other formats. Kartik worked on implementing this codec under the FFmpeg suite of libraries by taking reference from the FLIF specifications and reference codec.
Results: Successful
Student: Kartik K. Khullar (And Anamitra Ghurui)
Mentor: darkapex
Backup Mentor: Thilo Borgmann
JPEG 2000 Decoder improvements
Description:
FFmpeg contains a native JPEG 2000 decoder. It is missing many features so the expected result of this project is an JPEG 2000 decoder which can decode more samples.
Results: Successful
Student: Gautam Ramakrishnan
Mentor: Michael Niedermayer
Improving Color Constancy Video Filter
Description:
Description: Color Constancy is the ability to perceive colors of objects, invariant to the color of the light source. A Color Constancy video filter typically tries to remedy the effect of the scene illumination either by estimating it and then correcting the scene colors accordingly or directly mapping the scene colors to their most probable origin. Currently, FFmpeg has a basic variation that was developed in a previous GSoC project; but color constancy is an active field of research and a lot of new variations have been introduced that uses advanced image processing algorithms and machine learning.
Results: Not completed
Student: Yatendra
Mentor: Mina Sami
Backup Mentor: Thilo Borgmann
