Multi-view convolutional neural networks for video processing
Multi-view CNNs address the challenges of high-resolution video processing by encoding video data into multiple representations, enhancing efficiency and quality while reducing resource usage.
WO2025058768A1 Publication Date: 2025-03-20QUALCOMM INC
6 Cites 0 Cited by
Patent Information
- Application Number
- PCT/US2024/042089
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-09-13
- Filing Date
- 2024-08-13
- Publication Date
- 2025-03-20
AI Technical Summary
Technical Problem
Existing video processing technologies face challenges in efficiently handling high-resolution video data due to increased computational complexity and resource requirements, which can lead to suboptimal quality and increased energy consumption.
Method used
The use of multi-view convolutional neural networks (CNNs) for video processing, where input video data is partitioned into clips and encoded into multiple representations using a CNN encoder, allowing for efficient processing and decoding of video frames.
Benefits of technology
This approach improves the efficiency and quality of video processing by reducing computational resources and energy consumption while maintaining high visual quality, even with high-resolution video data.
✦ Generated by Eureka AI based on patent content.
Abstract
The present disclosure relates to processing video data. Some aspects involve partitioning input video data into two or more clips, each clip comprising a number of T frames, wherein each frame comprises a frame height, a frame width and a frame channel dimension Cin. Each clip is encoded into S encoded representations comprising a code height, a code width, and a code channel dimension, wherein T and S are integers with T ≥ S > 1. Encoding each clip into the S encoded representations may comprise concatenating all T frames of the clip into an input tensor along the frame channel dimension and encoding the input tensor into the S encoded representations using a convolutional neural network (CNN) encoder.
Need to check novelty before this filing date? Find Prior Art