toolcompass.

Find your next AI tool

Search by product name or task. Press Escape to close.

EoMT

● Identity checked

Images · Vendor site 2026-10-04

EoMT (Encoder-only Mask Transformer) is a CVPR 2025 research project that turns a plain Vision Transformer into a fast image segmentation model without adapters or decoders. The site publishes the paper overview, benchmarks, and links to open-source training code for semantic, instance, and panoptic segmentation.

Updated 2026-10-04View sources
Official website
Category
Images
Free access
Research code and the project page are publicly available with no paid subscription.
API access
Not confirmed

Is EoMT right for you?

A good fit for

Computer vision researchers and engineers experimenting with ViT-based segmentation who want the published EoMT implementation.

Before you choose

EoMT is a research release with training code rather than a hosted SaaS segmentation API on the project site.

ViT as a segmentation model

The EoMT project argues a plain Vision Transformer can serve as a minimalist, high-speed image segmentation model when pretrained ViTs are reused as the full architecture.

What it can do

Features & capabilities

Unknown is different from unavailable. Each fact carries its own evidence.

CapabilityValueEvidenceChecked
Segmentation approachOverview describes repurposing a plain ViT to encode image patches and segmentation queries as tokens without adapters or decoders.Facts sourced2026-10-04
Speed claimProject page states EoMT can be up to 4× faster than complex segmentation stacks when using ViT-L.Facts sourced2026-10-04
Code releaseHomepage links to GitHub repository tue-mps/eomt for implementation access.Facts sourced2026-10-04

Understand the total cost

EoMT pricing & plans

Free

Free

Research code and the project page are publicly available with no paid subscription.

Access
Research code and the project page are publicly available with no paid subscription.
Explore pricing & history

Alternatives to EoMT

View all ↗

The practical questions

Frequently asked questions

What is EoMT used for?

EoMT (Encoder-only Mask Transformer) is a CVPR 2025 research project that turns a plain Vision Transformer into a fast image segmentation model without adapters or decoders. The site publishes the paper overview, benchmarks, and links to open-source training code for semantic, instance, and panoptic segmentation.

Does EoMT have a free plan?

Research code and the project page are publicly available with no paid subscription.. This record lists ongoing free access; check the plan limits before starting.

How much does EoMT cost?

No paid monthly price is listed; this record treats the product as free to start. See the plan cards for entitlements, billing commitments and seat minimums.

Can I use EoMT through an API?

Not confirmed. API access and subscription access may have different terms; consult the linked sources.

What should I check before choosing it?

EoMT is a research release with training code rather than a hosted SaaS segmentation API on the project site.

Price history

No retained pricing changes yet. A current price alone does not establish a historical trend.

How this profile is supported

Facts apply to the named version and check date. Send a sourced correction if something changed.