Sapiens2

TL;DR

Sapiens2 introduces a versatile high-resolution transformer model family for human-centric vision tasks, achieving state-of-the-art results across multiple benchmarks through novel pretraining and architectural innovations.

Contribution

The paper presents Sapiens2, a new high-resolution transformer model with improved pretraining strategies, dataset, and architecture, enabling superior performance on diverse human-centric vision tasks.

Findings

Sapiens2 outperforms previous models on pose estimation (+4 mAP)

Achieves significant improvements in body-part segmentation (+24.3 mIoU)

Reduces normal estimation error by 45.6% in angular error

Abstract

We present Sapiens2, a model family of high-resolution transformers for human-centric vision focused on generalization, versatility, and high-fidelity outputs. Our model sizes range from 0.4 to 5 billion parameters, with native 1K resolution and hierarchical variants that support 4K. Sapiens2 substantially improves over its predecessor in both pretraining and post-training. First, to learn features that capture low-level details (for dense prediction) and high-level semantics (for zero-shot or few-label settings), we combine masked image reconstruction with self-distilled contrastive objectives. Our evaluations show that this unified pretraining objective is better suited for a wider range of downstream tasks. Second, along the data axis, we pretrain on a curated dataset of 1 billion high-quality human images and improve the quality and quantity of task annotations. Third, architecturally, we incorporate advances from frontier models that enable longer training schedules with improved stability. Our 4K models adopt windowed attention to reason over longer spatial context and are pretrained with 2K output resolution. Sapiens2 sets a new state-of-the-art and improves over the first generation on pose (+4 mAP), body-part segmentation (+24.3 mIoU), normal estimation (45.6% lower angular error) and extends to new tasks such as pointmap and albedo estimation. Code: https://github.com/facebookresearch/sapiens2