Qualcomm Unveils IMSDK 2.0: Bridging Generative AI and Multimedia to Accelerate Edge AI App Development

Perceive Beating AI News: Qualcomm has released the Qualcomm Intelligent Multimedia SDK (IMSDK) 2.0, unveiling a unified AI and multimedia development framework for the Qualcomm Dragonwing and Linux edge platforms. This framework aims to streamline the integration of camera, audio/video, sensor, AI inference, and cloud connectivity functions, accelerating the deployment of intelligent cameras, robots, drones, and industrial AI products.

The IMSDK 2.0 introduces new features such as Python and C++ Pipeline APIs, an Application Builder, generative AI inference, containerized microservices, AI programming agent skills, and "docs as code." Developers can leverage various inference runtimes such as QAIRT, ONNX Runtime, and TFLite to switch AI models without restructuring the entire multimedia pipeline.

Qualcomm stated that IMSDK 2.0 is built on the GStreamer architecture, incorporating hardware acceleration plugins, zero-copy data transfer, and multi-stream, multi-model pipeline capabilities to support visual AI, LLM/VLM, audio AI, and live text applications. Its containerized microservices span across Inference, Analytics, and Platform services, supporting enterprise-level connections such as Kafka, MQTT, AWS IoT, and Azure IoT.

Furthermore, IMSDK 2.0 introduces AI programming agent skills to assist developers in pipeline configuration, model integration, compilation, debugging, and deployment through natural language. Qualcomm mentioned that this SDK has been adopted by companies like Samsung, Amazon, and Bose for their product development.