Starstruk Logo MarkStarstruk

v0.60.0

roboflow/inferencev0.60.0Nov 7, 2025by grzegorz-roboflow

AI Summary

The release focuses on stream processing capabilities by adding RTMP ingestion and processing with WebRTC streaming back, alongside Jetson 6.2.0 build support. It introduces a `USE_INFERENCE_EXP_MODELS` flag for model selection and enables TensorRT (TRT) compilation within the CLI. Maintenance updates include fixes for bursty RTSP streams and dependency management.

Key Highlights

  • RTMP stream support (ingest, process, stream back)
  • Jetson 6.2.0 build support
  • TRT compilation in `inference-cli`
  • `USE_INFERENCE_EXP_MODELS` flag

New Features

  • RTMP Streaming
  • Jetson 6.2.0
  • TRT CLI Compilation
  • USE_INFERENCE_EXP_MODELS

Full Release Notes

## What's Changed
* Serialize parent origin metadata for crops/slices by @brunopicinin in https://github.com/roboflow/inference/pull/1672
* Update perspective correction for dimensionality by @lou-roboflow in https://github.com/roboflow/inference/pull/1674
* "USE_INFERENCE_EXP_MODELS" by @hansent in https://github.com/roboflow/inference/pull/1657
* Add support for ingesting and processing rtmp stream remotely (on modal or process) and stream back processed frames over webrtc by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1675
* Add GitHub Actions workflow for Jetson 6.2.0 builds by @alexnorell in https://github.com/roboflow/inference/pull/1671
* uv lock inference-experimental by @hansent in https://github.com/roboflow/inference/pull/1677
* Add rtsp placeholder to webrtc_worker by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1683
* Set HF_HOME when running webrtc_worker on Modal by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1682
* Make webrtc_worker time out gracefully if WEBRTC_MODAL_FUNCTION_TIME_LIMIT is set by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1684
* add file lock for downloading core models by @hansent in https://github.com/roboflow/inference/pull/1680
* TRT Compilation of models in `inference-cli` by @PawelPeczek-Roboflow in https://github.com/roboflow/inference/pull/1679
* Remove (currently broken) `trt-*-cu13` dependencies from `inference-exp` by @PawelPeczek-Roboflow in https://github.com/roboflow/inference/pull/1689
* Handle bursty rtsp by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1691
* Optimize bytes_to_unicode by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1676
* Drain rtsp queue by @grzegorz-roboflow in https://github.com/roboflow/inference/pull/1693
* InferenceExpObjectDetectionModelAdapter: dont optimize_for_inference  by @hansent in https://github.com/roboflow/inference/pull/1694


**Full Changelog**: https://github.com/roboflow/inference/compare/v0.59.1...v0.60.0