Search papers, labs, and topics across Lattice.
This paper extends a prior VTA compiler for FPGA-based CNN acceleration, overcoming limitations in on-chip memory capacity and automation. The extended compiler now supports the compilation of complete CNNs, enabling the processing of larger models. The effectiveness of the enhanced compilation chain is validated through the successful compilation and simulated execution of a YOLO-NAS object detection model on the VTA.
Compiling and executing YOLO-NAS on an FPGA-based accelerator is now possible, opening doors for real-time object detection in safety-critical applications like aeronautics.
Deploying complex Convolutional Neural Networks (CNNs) on FPGA-based accelerators is a promising way forward for safety-critical domains such as aeronautics. In a previous work, we have explored the Versatile Tensor Accelerator (VTA) and showed its suitability for avionic applications. For that, we developed an initial stand-alone compiler designed with certification in mind. However, this compiler still suffers from some limitations that are overcome in this paper. The contributions consist in extending and fully automating the VTA compilation chain to allow complete CNN compilation and support larger CNNs (which parameters do not fit in the on-chip memory). The effectiveness is demonstrated by the successful compilation and simulated execution of a YOLO-NAS object detection model.
