Exposing Video Compression History by Detecting Transcoded HEVC Videos from AVC Coding

Abstract

The analysis of video compression history is one of the important issues in video forensics. It can assist forensics analysts in many ways, e.g., to determine whether a video is original or potentially tampered with, or to evaluate the real quality of a re-encoded video, etc. In the existing literature, however, there are very few works targeting videos in HEVC format (the most recent standard), especially for the issue of the detection of transcoded videos. In this paper, we propose a novel method based on the statistics of Prediction Units (PUs) to detect transcoded HEVC videos from AVC format. According to the analysis of the footprints of HEVC videos, the frequencies of PUs (whether in symmetric patterns or not) are distinguishable between original HEVC videos and transcoded ones. The reason is that previous AVC encoding disturbs the PU partition scheme of HEVC. Based on this observation, a 5D and a 25D feature set are extracted from I frames and P frames, respectively, and are combined to form the proposed 30D feature set, which is finally fed to an SVM classifier. To validate the proposed method, extensive experiments are conducted on a dataset consisting of CIF (352 × 288) and HD 720p videos with a diversity of bitrates and different encoding parameters. Experimental results show that the proposed method is very effective at detecting transcoded HEVC videos and outperforms the most recent work.

Double JPEG Compression Detection Based on Noise-Free DCT Coefficients Mixture Histogram Model

“…Besides, some novel image segmentation technologies can be considered for locating the tampered region more precisely and accelerating our method. Finally, the double compression detection tasks on some other media such as audio [28] and video [29,30] will be considered in the future.…”

A Review on HEVC Video Forensic Investigation under Compressed Domain

“…In the entertainment sector, where hundreds of hours of videos are uploaded to YouTube every minute, quality control measures must be extremely efficient. In [10,20,21], the authors presented forensic approaches for determining if it is an authentic HEVC video or it is generated by transcoding an existing MPEG/AVC video sequence.…”

“…Bian et al [20] explored the differences in prediction unit statistics of I-frame and P-frames for single and double encoded videos. Differences in quantization parameters between the first and second encoding cause the distortion of the optimal encoding scheme thus, to achieve the least cost the partitioning scheme of transcoded video tries different patterns.…”

Detection of Transcoding from H.264/AVC to HEVC Based on CU and PU Partition Types

“…However, it is only suitable for videos encoded with constant Quantization Parameter (QP) mode, and it needs to recompress the videos a number of times, which leads to high computational complexity. Bian et al [15] combined the frequency of PU partition types in I pictures and P pictures to detect AVC/HEVC videos, and the results confirmed that PU partition types could be effective in identifying AVC/HEVC videos. However, it did not consider the characteristic of CU partition types.…”

“…However, it did not consider the characteristic of CU partition types. Like Reference [15], Reference [12], which was proposed for detecting videos recompressed with the same codec as the original one, also only explored PU partition types and did not consider the CU partition types. Furthermore, PU partition types were only extracted from P pictures in Reference [12].…”