v0.64.8

roboflow/inferencev0.64.8Feb 13, 2026by grzegorz-roboflow

AI Summary

Feature release adding fisheye camera calibration support and a heatmap visualization block for detections and tracklets. Includes various performance optimizations and maintenance improvements.

Key Highlights

  • Fisheye cameras support in camera calibration block
  • Heatmap block for detections and tracklets
  • Enable gzip compression by default
  • Fix video frame count estimation from uploaded video
  • Add frame size to WebRTC video metadata
  • SAM3 visual segmentation enabled

New Features

  • Fisheye camera calibration
  • Heatmap visualization block (supports detections and tracklets with movement tracking)

Full Release Notes

# 💪 Added

* Fisheye cameras in camera calibration block by @Erol444 in https://github.com/roboflow/inference/pull/1996
Calibration block was supporting polynomial calibration which is not handling fisheye distortions well. This change adds support for fisheye calibration.
<img width="1607" height="805" alt="image" src="https://github.com/user-attachments/assets/cb37deec-6226-4a33-9765-70720afd1267" />

* Heatmap block by @Erol444 in https://github.com/roboflow/inference/pull/1986
This change adds heatmap block (uses supervision's heatmap annotator), which supports both:
 - detections, so heatmap based on where detections were
 - tracklets, which ignores stationary objects (default: on), so we heatmap the movements not the objects

https://github.com/user-attachments/assets/6106ace7-2d8d-4ebd-ab42-d89f7af8121d

# 🚧 Maintanence
* temporarily pin z3-solver version by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1990
* Code workflow block icon issue by @Erol444 in https://github.com/roboflow/inference/pull/1988
* Optimize cosine_similarity by @KRRT7 in https://github.com/roboflow/inference/pull/1989
* add inference version to the request headers by @japrescott in https://github.com/roboflow/inference/pull/1985
* Fix video frame count estimation by detecting actual FPS from uploaded video by @rafel-roboflow in https://github.com/roboflow/inference/pull/1992
* Mark file processing in webrtc worker for downstream blocks to pick frame timestamp correctly by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1995
* add frame size to webrtc video metadata by @rafel-roboflow in https://github.com/roboflow/inference/pull/1997
* enable gzip compression by default by @rafel-roboflow in https://github.com/roboflow/inference/pull/1998
* WIP: enabled sam3 visual segment by @rafel-roboflow in https://github.com/roboflow/inference/pull/1975
* added ffmpeg to docker dependencies by @rafel-roboflow in https://github.com/roboflow/inference/pull/2002
* rename seg-preview to sam3 by @rafel-roboflow in https://github.com/roboflow/inference/pull/2005
* Fix RF-DETR-Seg mask postprocessing for letterboxed input case by @mkaic in https://github.com/roboflow/inference/pull/2001
* Enable inference pipeline api on jetpack 6.2.0 by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/2006


**Full Changelog**: https://github.com/roboflow/inference/compare/v0.64.7...v0.64.8