Unsupervised Training for 3D Morphable Model Regression
Citations Over TimeTop 1% of 2018 papers
Abstract
We present a method for training a regression network from image pixels to 3D morphable model coordinates using only unlabeled photographs. The training loss is based on features from a facial recognition network, computed on-the-fly by rendering the predicted faces with a differentiable renderer. To make training from features feasible and avoid network fooling effects, we introduce three objectives: a batch distribution loss that encourages the output distribution to match the distribution of the morphable model, a loopback loss that ensures the network can correctly reinterpret its own output, and a multi-view identity loss that compares the features of the predicted 3D face and the input photograph from multiple viewing angles. We train a regression network using these objectives, a set of unlabeled photographs, and the morphable model itself, and demonstrate state-of-the-art results.
Related Papers
- → A Characterization for Generalized Hukuhara Differentiable Interval-Valued Functions and Some Rules of Calculus(2022)2 cited
- → Continuous Differentiability in the Context of Generalized Approach to Differentiability(2023)1 cited
- → The Complexity Of Nowhere Differentiable Continuous Functions(1989)1 cited
- On binary function continuity,differentiability and may differentiability(2012)
- → On derivatives of fuzzy multi-dimensional mappings and applications under generalized differentiability(2021)