GStreamer Conference 2019
The 10th GStreamer Conference took place on Thursday and Friday 31 October - 1 November 2019 in Lyon, France. It was attended by about 200 participants who were developers, community members, decision-makers, industry partners, or researchers in the GStreamer multimedia framework. Intel presented the GStreamer Video Analytics (GVA) plugin developed by Intel to support inference using the Intel® Distribution of OpenVINO™ Toolkit. The presentation provided background on growing demand for video analytics, typical video analytics pipeline, Intel hardware for analytics, and OpenVINO™ Toolkit support for optimized deep learning inference across all the Intel hardware. Then, it focused on GStreamer plugin for Video Analytics which uses OpenVINO™ Inference Engine and provided some video analytics pipeline examples with live demos such as face detection and classification, multi-channel object detection, and smart cities. The presentation and demos aroused audience interest about applications and use cases of video analytics, pretrained models for inference, model training, and production readiness of Intel technologies. Intel’s GStreamer Video Analytics (GVA) plugin is an open source project. The beta version available on GitHub, includes some sample pipeline scripts, a user guide, and API reference documentation. To listen to the talk at GStreamer conference, please click on the link below. Please visit our GVA plugin GitHub repo, then download and start building your video analytics solutions.
- GStreamer conference talk: https://gstconf.ubicast.tv/videos/gstreamer-video-analytics-optimizing-inference-across-hw-targets-neelay-shah-neena-maldikar-mikhail-nikolsky-ilya-belyakov-intel/
- GStreamer Conference 2019 site: https://gstreamer.freedesktop.org/conference/2019/
- Github repo for GVA plugin: https://github.com/opencv/gst-video-analytics
- OpenVINO Toolkit: https://software.intel.com/en-us/openvino-toolkit