AI Benchmark

AI Benchmark is a tool designed to evaluate the AI performance of Android smartphones, specifically their capability to run the latest Deep Neural Networks for various AI-based tasks. It provides a professional assessment of speed, accuracy, power consumption, and memory requirements for key AI, Computer Vision, and NLP models. The solutions tested include Image Classification, Face Recognition methods, AI models for neural image and text generation, neural networks for Image/Video Super-Resolution and Photo Enhancement, and AI solutions used in autonomous driving systems and smartphones for real-time Depth Estimation and Semantic Image Segmentation. The tool also offers graphical visualization of the algorithms’ outputs, providing an understanding of the current state-of-the-art in various AI fields.

In addition, AI Benchmark Alpha is an open-source python library that evaluates the AI performance of different hardware platforms, including CPUs, GPUs, and TPUs. It utilizes the TensorFlow machine learning library, offering a precise and lightweight solution for assessing inference and training speed for key Deep Learning models. AI Benchmark can be downloaded as a Python pip package to any system running Windows, Linux, or macOS.

The comprehensive tests cover all major Deep Learning tasks and architectures, making it a useful tool for researchers, developers, hardware vendors, and end-users who run AI applications on their devices. Detailed setup information for each test, including input and batch sizes, test modes, can be found on the ranking page.