GoAI 2.0| GOWIN Semiconductor

GoAI 2.0 is a machine learning development platform to deploy CNNs (Convolutional Neural Networks) on GOWIN FPGAs. It includes software scripts nedded to convert TensorFlow files and accelerator FPGA IP to run trained machine learning models as well as tutorials and examples.

GoAI 2.0 is integrated into the TensorFlow and TensorFlow Lite framework providing superior ease of use over legacy and alternative platforms.

GoAI 2.0 extends upon TensorFlow's extensive software development environment by generating all the necessary firmware to deploy a machine learning model directly onto GOWIN FPGAs without writing any additional RTL or C code. Windows and Linux including Ubuntu are fully supported.

Source
		GoAI Development Kits GW1NSR4 GW2AR18 GW2A55

		How to train a model on Google Cloud Platform Car Detection Model Training [Video] How to generate a flatbuffers (.tflite) file in Google Colab Using Google Colab [Video] How to generate FPGA/MCU Firmware files from a flatbuffers (.tflite) file Video How to run SDK and deploy onto GoAI Development Board Using GUI [Video] (coming soon) Using command line [Video] (coming soon)

		Camera Input Designs for Object Detection FPGA and MCU Firmware Files: GW2A55 GW2AR18 [Coming Soon] GW1NSR4 Coefficient Files: Person
		HDMI/DVI Input Designs for Object detection FPGA and MCU Firmware Files: GW2A55 [Coming Soon] Coefficient Files: Car Detection [Coming Soon]
		Audio Input Designs for Speech and Sound Detection FPGA and MCU Firmware Files: GW2A55 GW2AR18 [Coming Soon] GW1NSR4 [Coming Soon]
		Accelerometer for Gesture and Movement Detection [Coming Soon] FPGA and MCU Firmware Files: GW2A55 [Coming Soon] GW2AR18 [Coming Soon] GW1NSR4 [Coming Soon] Coefficient Files: 'W'/'O'/'<' Human Phrases [Coming Soon]
		Data Input Prediction FPGA and MCU Firmware Files: GW2A55 GW2AR18 GW1NSR4

		Useful Links GoAI 2.0 Developer FAQ Netron - flatbuffers (*.tflite visualizer GUI)