PolyMaX: General Dense Prediction with Mask Transformer

TL;DR

PolyMaX introduces a unified mask transformer framework for various dense prediction tasks, achieving state-of-the-art results by generalizing cluster-prediction methods to continuous outputs like depth and surface normals.

Contribution

The paper generalizes mask transformer-based cluster prediction to all dense prediction tasks, unifying discrete and continuous output predictions within a single framework.

Findings

State-of-the-art performance on NYUD-v2 benchmarks

Effective unification of dense prediction tasks

Inspired by success of discretization in depth estimation

Abstract

Dense prediction tasks, such as semantic segmentation, depth estimation, and surface normal prediction, can be easily formulated as per-pixel classification (discrete outputs) or regression (continuous outputs). This per-pixel prediction paradigm has remained popular due to the prevalence of fully convolutional networks. However, on the recent frontier of segmentation task, the community has been witnessing a shift of paradigm from per-pixel prediction to cluster-prediction with the emergence of transformer architectures, particularly the mask transformers, which directly predicts a label for a mask instead of a pixel. Despite this shift, methods based on the per-pixel prediction paradigm still dominate the benchmarks on the other dense prediction tasks that require continuous outputs, such as depth estimation and surface normal prediction. Motivated by the success of DORN and AdaBins in depth estimation, achieved by discretizing the continuous output space, we propose to generalize the cluster-prediction based method to general dense prediction tasks. This allows us to unify dense prediction tasks with the mask transformer framework. Remarkably, the resulting model PolyMaX demonstrates state-of-the-art performance on three benchmarks of NYUD-v2 dataset. We hope our simple yet effective design can inspire more research on exploiting mask transformers for more dense prediction tasks. Code and model will be made available.