Falcon Perception

Falcon Perception is an early-fusion vision-language model from the Technology Innovation Institute. In BOAT it provides text-prompted detection and segmentation on every supported platform, including CPU-only Linux machines.

Object DetectionInstance SegmentationSemantic SegmentationTIIApache-2.0

Recommended2/10

VRAM2.5 GBweights, plus ~0.5 GB reserve

Tools2across 3 annotation types

Runs onmacOSCUDACPU4 of 4 runtime profiles

Install

boat model install falcon-perception

Installs falcon-perception:600m with 2.5 GB of weights into an isolated runtime.

Tools

What Falcon Perception adds to the annotation editor. Each example loops what you do and what comes back.

Text to Mask

Text promptMasks

Type a class name such as “apple” and press Enter. BOAT previews every matching mask before you apply them.

Preview, then apply

Text to Boxes

Text promptBounding boxes

Object Detection

Type a class name such as “car” and press Enter. BOAT previews every matching bounding box before you apply them.

Preview, then apply

Tag	Weights	Platforms	Install
falcon-perception:600mdefault	2.5 GB · 6 files	macOSCUDACPU

Platforms

BOAT picks the runtime profile that matches your machine. Only that profile's code is downloaded.

macOS · Metal

Apple silicon (MPS)

Python 3.11 · 58 KB

Linux · CUDA 12.8

NVIDIA, incl. Blackwell

Python 3.11 · 61 KB

Linux · CUDA 12.6

NVIDIA GPUs

Python 3.11 · 62 KB

Linux · CPU

x86_64 CPU

Python 3.11 · 58 KB

License

BOAT's adapter code is Apache-2.0. The model's own code and weights keep their upstream license:Apache-2.0.

Bundled license files: BOAT-APACHE-2.0.txt, NOTICE.md, UPSTREAM-LICENSE.txt

Package details

Package: falcon-perception
Version: 0.1.0
Type: Common model
BOAT Core: ≥ 0.1.0 · < 1.0.0
Idle unload: 300s (change with boat model setup)
Inference timeout: 1800s
Manifest SHA-256: 8c31a00a8ccd4df946cd52e99309894562f3becd5b35c32049377e0e69d6d70c
Manifest: manifests/falcon-perception/0.1.0-8c31a00a8ccd4df9.json

Upstream notice

BOAT adapter code is Apache-2.0. Upstream model code and weights keep their own licenses.

Model: tiiuae/Falcon-Perception
Pinned weight provenance is in ../weights.json. Vendored code revisions, when present,
are recorded in ../vendor/UPSTREAM_REVISION. The local macOS adaptations are retained.
Runtime dependencies are installed from pinned distributions with their own license files.

Upstream license: Apache-2.0 (model card `license: apache-2.0`)
Model card: https://huggingface.co/tiiuae/Falcon-Perception