Super-Resolution Image Generation Circuit With Dynamic Block Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional SR image generation circuits lack flexibility in supporting different depth structures and/or different numbers of channels due to fixed early-stage determinations of basic block configurations.
Innovation Solution
Incorporation of a configurable basic block pool circuit that dynamically configures basic blocks according to input image depth requirements and color encoding format channels, allowing for adaptable SR image generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the depth of basic blocks and number of sets of basic blocks are determined at an early stage in hardware design, then the circuit structure is fixed and simple to implement, but the design flexibility and adaptability to different depth structures and channel numbers are limited
Solution Approach 1:
The patent implements dynamic configuration of basic blocks through a configurable basic block pool circuit. The depth of basic blocks and number of sets of basic blocks are no longer fixed at hardware design stage but can be dynamically adjusted according to different processing requirements. This allows the same hardware circuit to adapt to various depth structures and channel numbers by reconfiguring the basic block pool, thereby resolving the contradiction between adaptability and complexity.
Solution Approach 2:
The configurable basic block pool circuit serves multiple functions: it can be configured to provide different depths of basic blocks for both stable SR processing and GAN processing circuits, and can also be configured to provide different numbers of sets of basic blocks according to color encoding format channel requirements. This universal configuration mechanism eliminates the need for separate fixed circuits for each configuration, reducing overall device complexity while enhancing adaptability.
2Adaptability or versatility
If multiple fixed configurations of basic blocks are provided to support different depth structures and channel numbers, then adaptability is improved, but device complexity and hardware resources increase
Solution Approach 1:
The patent merges the configuration functions for depth adjustment and channel number adjustment into a single configurable basic block pool circuit. Instead of providing separate fixed circuits for each possible configuration, the basic block pool is combined into one reusable resource that can be dynamically allocated to both stable SR processing and GAN processing circuits. This merging approach reduces hardware resources while maintaining support for different depth structures and channel numbers.
Solution Approach 2:
The basic block pool implements dynamic allocation where the same physical hardware resources can be reconfigured to serve different logical configurations. By making the basic block pool dynamically configurable rather than providing multiple fixed instances, the circuit supports various depth structures and channel numbers using the same underlying hardware, thereby reducing overall device complexity and resource usage.
Data Source
AI summary
A super resolution (SR) image generation circuit includes an image scale-up circuit, a stable SR processing circuit, a generative adversarial network (GAN) processing circuit, and a configurable basic block pool circuit. The image scale-up circuit is arranged to receive and process an input image to generate a scaled-up image. The stable SR processing circuit is arranged to receive a feature map of the input image to generate a stable delta value. The GAN processing circuit is arranged to receive the feature map to generate a GAN delta value. The configurable basic block pool circuit is arranged to dynamically configure a plurality of basic blocks according to a depth requirement of the input image, to generate a configuration result. The SR image generation circuit generates an SR image according to the scaled-up image, the stable delta value, and the GAN delta value.