| 102 | | |
| 103 | | == Super Resolution filter == |
| 104 | | |
| 105 | | '''Description:''' Use machine learning super resolution techniques to implement an upscaling filter. The student must evaluate the best suited method based on state of the art methods |
| 106 | | |
| 107 | | '''Expected results:''' A working super resolution filter, which could be trained for specific content. |
| 108 | | |
| 109 | | '''Prerequisites:''' Good C coding skills, knowledge of machine learning and convolutional neural networks, basic familiarity with git, familiarity with one machine learning framework. |
| 110 | | |
| 111 | | '''Qualification Task:''' Implement a ffmpeg filter which performs the [https://arxiv.org/abs/1501.00092 SRCNN] 2x upscale, pre-trained weights will be provided. |
| 112 | | |
| 113 | | '''Wiki page:''' [[SponsoringPrograms/GSoC/2018/SR|Super Resolution]] |
| 114 | | |
| 115 | | '''Mentor:''' Pedro Souza (bygrandao [at] gmail [dot] com) |
| 116 | | |
| 117 | | '''Backup Mentor:''' Michael Niedermayer (michaelni@gmx.at) |
| 118 | | |
