Authors
Michael Fischer, Tobias Ritschel
University College London
Portals
Abstract
There currently exist two main approaches to reproducing visual appearance using Machine Learning (ML): The first is training models that generalize over different instances of a problem, e.g., different images of a dataset. As one-shot approaches, these offer fast inference, but often fall short in quality. The second approach does not train models that generalize across tasks, but rather over-fit a single instance of a problem, e.g., a flash image of a material. These methods offer high quality, but take long to train. We suggest to combine both techniques end-to-end using meta-learning: We over-fit onto a single problem instance in an inner loop, while also learning how to do so efficiently in an outer-loop across many exemplars. To this end, we derive the required formalism that allows applying meta-learning to a wide range of visual appearance reproduction problems: textures, BRDFs, svBRDFs, illumination or the entire light transport of a scene. The effects of meta-learning parameters on several different aspects of visual appearance are analyzed in our framework, and specific guidance for different tasks is provided. Metappearance enables visual quality that is similar to over-fit approaches in only a fraction of their runtime while keeping the adaptivity of general models.
Contribution
- Metappearance1 , a model that adapts to new, unseen visual appearance tasks in only a few steps of gradient descent
- Optimizing for a fast and accurate optimizer of this model
- Instances of this model that accurately match texture, BRDFs, svBRDFs, illumination, or light transport orders of magnitude faster than strong baselines, at comparable quality, an
- An analysis of our method’s properties, its convergence and its behaviour under ablation
Related Works
Visual Appearance; Learning; General Learning; Over-fit Optimization; Fine-tuning; Hyper- and meta-learning