Lakshantha Dissanayake

@johnnynunez got it! Thanks! These can be integrated to our CI's at the end of year then :)

Hi @johnnynunez. Great news! Thanks for the share :) I will implement this in the following days!

Hello @harshdhamecha, Thank you for your interest in RKNN integration. We are currently in the development phase of this integration and have made significant progress. At this stage, we have...

Yes @yg-dickson . Once the integration is finished, it will support YOLO11.

@yg-dickson INT8 export of RKNN is not yet supported. We will work on it. By default, it should be FP16.

> @lakshanthad works with `yolo11`. I will solve the conflicts asap. Great! Thanks @ambitious-octopus . I will also verify from my side by running on the Rockchip device.

> @lakshanthad also work with the latest `rknn-toolkit2` version `2.3.0` @ambitious-octopus sounds good. Will also check this on the Rockchip device side.

@ambitious-octopus YOLO11 export on Linux (with `rknn-toolkit2` version `2.3.0`) + YOLO11 inference on Rockchip device (with `rknn-toolkit-lite2` version `2.3.0`) works! I have done some minor fixes as well. Waiting for...

@ambitious-octopus I have reviewed everything and done further optimizations. What's remaining: 1. Main blog image - Waiting for Sony to confirm this. Probably will use: https://www.raspberrypi.com/documentation/accessories/images/ai-camera.png?hash=0b11215f8c0b935379f97f5bd47910f1 2. Benchmarks - We...

Hey guys, I can confirm that after reflashing with JetPack6.0 which includes TensorRT 8.6.2, INT8 engine generation works as expected. The issue is only there when using JetPack6.1 which includes...