Interstitial 3D Scene Data for Video Authenticity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Deep fake technology poses a significant threat to the authenticity of video recordings, making it difficult to distinguish manipulated videos from original ones, which can have serious implications in various fields such as politics and criminal law, as existing digital signatures can be deceived by altering videos on a computer and then recording them from a screen.
Innovation Solution
Incorporating interstitial three-dimensional (3D) scene information into video streams using a digital video camera equipped with a 3D scanner and compiling logic, which inserts this data into the video stream, allowing for the creation of a real video signature that authenticates the video by verifying the physical setting of the recording.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If video compression techniques using keyframes and inter-frame prediction are used, then video data transmission efficiency is improved, but video authenticity verification capability deteriorates
Solution Approach 1:
The patent segments video authentication into two parts: (1) capturing 3D scene information at specific intervals (keyframes) rather than continuously, and (2) inserting authentication data only at those segmented points. This allows efficient compression while maintaining verification capability at critical moments.
Solution Approach 2:
The patent introduces an intermediary authentication mechanism that bridges compressed video frames. By embedding 3D scene information and cryptographic signatures as intermediary data structures within the video stream, the system enables authenticity verification without requiring full uncompressed video data.
2Reliability
If 3D scene information is inserted into every video frame, then video authenticity verification is improved, but video stream complexity and data volume increase
Solution Approach 1:
The patent implements periodic insertion of 3D scene information and authentication data at regular intervals (e.g., every keyframe or every N frames) rather than continuously. This periodic authentication approach maintains verification capability while significantly reducing stream complexity and data volume.
Solution Approach 2:
The patent applies different quality levels of authentication to different parts of the video stream. Full 3D scene information is captured at keyframe intervals, while intermediate frames use lighter authentication mechanisms or reference previous authentication points, creating local variations in authentication intensity.
3Reliability
If deep fake detection using 3D scene matching is implemented, then video authenticity is improved, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary capture of 3D scene information at the time of video recording, storing it as reference data within the video stream. During authentication, this pre-captured 3D data is directly compared with the video content, eliminating the need for time-consuming real-time 3D scanning and processing.
Solution Approach 2:
The patent creates a copy of the 3D scene information as it existed at the time of recording and embeds this copy within the video stream. Authentication involves comparing this embedded 3D copy with the video frames, which is computationally more efficient than performing new 3D scans during verification.
Data Source
AI summary
There is disclosed in one example a digital video camera, including: an analog picture element; an analog-to-digital converter (ADC) to digitize input from the analog picture element; a three-dimensional (3D) scanner; compiling logic to compile the digitized input into a video stream; and insertion logic to insert interstitial 3D scene data into the video stream.