SPAR: Single-Pass Any-Resolution ViT for Open-vocabulary Segmentation

Abstract

Foundational Vision Transformers (ViTs) have limited effectiveness in tasks requiring fine-grained spatial understanding, due to their fixed pre-training resolution and inherently coarse patch-level representations. These challenges are especially pronounced in dense prediction scenarios, such as open-vocabulary segmentation with ViT-based vision-language models, where high-resolution inputs are essential for accurate pixel-level reasoning. Existing approaches typically process large-resolution images using a sliding-window strategy at the pre-training resolution. While this improves accuracy through finer strides, it comes at a significant computational cost. We introduce SPAR: Single-Pass Any-Resolution ViT, a resolution-agnostic dense feature extractor designed for efficient high-resolution inference. We distill the spatial reasoning capabilities of a finely-strided, sliding-window teacher into a single-pass student using a feature regression loss, without requiring architectural changes or pixel-level supervision. Applied to open-vocabulary segmentation, SPAR improves single-pass baselines by up to 10.5 mIoU and even surpasses the teacher, demonstrating effectiveness in efficient, high-resolution reasoning. Code: https://github.com/naomikombol/SPAR

Workflow Status

Review status

pending

Role

unreviewed

Read priority

later

Vote

Not set.

Saved

no

Collections

Not filed yet.

Next action

Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

Similar Papers

BibTeX

@article{kombol2026spar,
  title = {SPAR: Single-Pass Any-Resolution ViT for Open-vocabulary Segmentation},
  author = {Naomi Kombol and Ivan Martinović and Siniša Šegvić and Giorgos Tolias},
  year = {2026},
  abstract = {Foundational Vision Transformers (ViTs) have limited effectiveness in tasks requiring fine-grained spatial understanding, due to their fixed pre-training resolution and inherently coarse patch-level representations. These challenges are especially pronounced in dense prediction scenarios, such as open-vocabulary segmentation with ViT-based vision-language models, where high-resolution inputs are essential for accurate pixel-level reasoning. Existing approaches typically process large-resolution },
  url = {https://arxiv.org/abs/2604.02252},
  keywords = {cs.CV},
  eprint = {2604.02252},
  archiveprefix = {arXiv},
}

Metadata

{}