Similar to how humans use their sense of sight and hearing as part of their cognitive process of decision making, at Esse Pi they apply different AI techniques to both Image/Video and Audio data streams to auto detect patterns and achieve the best prediction outcome.
Besides their public cloud-based platform, they also offer a secured on-premise platform to ensure that sensitive data never leaves the network. Esse Pi’s flagship CAP (Content Analysis Platform) provides a selection of customisable Machine Learning (ML) models and inference engines to contextualise media content. They uses text, audio and visual input to auto-detect scenes of violence, obscenity and profanity in the media content. The platform also provides built-in tools that enable annotation, dataset augmentation as well as self-training and management of custom ML models.
Esse Pi also provides AI at the Edge solutions that enable optimized AI algorithms to run directly on IoT edge device; this allows IoT applications to run in environment with limited internet connectivity or without incurring high mobile data cost. They offer an ARM-based IoT gateway that can be connected wirelessly and securely to a network of IoT sensors capturing multimodal inputs (image, audio and vibration). The IoT gateway will process the sensory data and run the optimised ML inference locally.