Megareus

Under Construction Project, Open to issue / PR submission

Megareus: A collection of experimental results of various image / video / point cloud processing algorithms

Abstract

There are more and more datasets and models released nowadays. From the aspects of researching, the exhaustive comparison between models for various data is a hard, complicated task. From the aspects of application, it's a problem to select a good model from the numerous models / algorithms. Megareus provides a collection of experimental results of open image / video processing algorithms with various published datasets and some of open license data.

Main Content

There are several kinds of algorithms included in this collection:

Test Content: Assassin's Creed Odyssey

Source:

https://youtu.be/2QyIP8Cf1HE

Preview:

• Human Face Detection

  Name of Algorithm / Model	Results
  Face Detection and Tracking Using the KLT Algorithm
  MTCNN Face Detection	Full result: https://youtu.be/OjOfELSAvss

  2. FDDB: Face Detection Data Set and Benchmark

• No-Reference Image Quality Assessment (NR-IQA)

  Name of Algorithm / Model	Results
  NIQE, PIQE, BRISQUE	Full result: https://youtu.be/ed07n1c49pQ

• Object Detection

• DCT(Discrete Cosine Transform) related experiments

  Please check the webpage "DCT.md".

Project Structure

References:

• Human Face Detection

  1. [Matlab File Exchange] Face Detection and Tracking Using the KLT Algorithm

  2. [Matlab File Exchange] MTCNN Face Detection

• Point Cloud Quality Assessment

  1. Yipeng Liu, Qi Yang, Yiling Xu, Le Yang, "Point Cloud Quality Assessment: Dataset Construction and Learning-based No-Reference Metric", ACM Transactions on Multimedia Computing Communications and Applications, 2022."

     Github repo: ResSCNN