FocsleFocsle
ME

Mask2Former-Mobile

4%
by Meta FAIR-Edge

Mobile-class panoptic segmenter. The reference for surgical, factory and medical pipelines.

SegmentationMITINT8FP16panoptictransformer
64K downloads 3.1K deploymentsUpdated Jan 21, 2028
Headline:12.4ms · NVIDIA Jetson Thor · FP16

About this model

Mobile-class panoptic segmenter. The reference for surgical, factory and medical pipelines.

Authored by meta-fair-edge. Curated into the Fo’c’sle reference set on 2028-01-21. All cross-chip benchmarks below were collected in matched-pair runs in the HIL lab using the same input pipeline, same upstream preprocessing, and the same downstream consumer. See the methodology page for the full protocol.

Task
Segmentation
Parameters
47.2 M
Benchmarked on
5 chips
Deployments
3.1K

Architecture

Mask transformer architecture
Inferred from upstream weights · simplified
ImageBackbonePixel decoderMask transformerClass queriesMask + class

Headline benchmarks

Training data

Pretrained on the upstream maintainer’s released checkpoint. Edge-distillation pass uses 2.4M frames from the Fo’c’sle distillation corpus (consented public data + opt-in publisher contributions). Quantization-aware fine-tune uses 320K calibration samples drawn from the target task’s eval domain.

  • Pretraining corpus: upstream maintainer release
  • Distillation corpus: 2,400,000 frames
  • Calibration set: 320,000 samples (per task)
  • Eval set: standard benchmark + matched-pair HIL runs