Learning Instance-Specific Parameters of Black-Box Models Using Differentiable Surrogates
This addresses the challenge of optimizing black-box models for specific inputs, which is incremental but useful for applications like image denoising.
The paper tackles the problem of tuning input-specific parameters for non-differentiable black-box models, such as BM3D for image denoising, by using differentiable surrogate models and neural networks, resulting in significant improvements in PSNR and SSIM nearing 0.93 on datasets like SIDD and CBSD68.
Tuning parameters of a non-differentiable or black-box compute is challenging. Existing methods rely mostly on random sampling or grid sampling from the parameter space. Further, with all the current methods, it is not possible to supply any input specific parameters to the black-box. To the best of our knowledge, for the first time, we are able to learn input-specific parameters for a black box in this work. As a test application, we choose a popular image denoising method BM3D as our black-box compute. Then, we use a differentiable surrogate model (a neural network) to approximate the black-box behaviour. Next, another neural network is used in an end-to-end fashion to learn input instance-specific parameters for the black-box. Motivated by prior advances in surrogate-based optimization, we applied our method to the Smartphone Image Denoising Dataset (SIDD) and the Color Berkeley Segmentation Dataset (CBSD68) for image denoising. The results are compelling, demonstrating a significant increase in PSNR and a notable improvement in SSIM nearing 0.93. Experimental results underscore the effectiveness of our approach in achieving substantial improvements in both model performance and optimization efficiency. For code and implementation details, please refer to our GitHub repository: https://github.com/arnisha-k/instance-specific-param