JetPack-3.3 for TX2
NVIDIA formally released JetPack-3.3 around end of July. According to the ReleaseNotes, the highlight of this 3.3 release is TensorRT 4.0 and its corresponding cuDNN libraries.
I tested/verified most of my previous posts with this JetPack-3.3. And I’m happy to report that all Caffe/TensorFlow models work as expected. More specifically, I’ve just verified the following on JetPack-3.3:
- How to Install OpenCV (3.4.0) on Jetson TX2 -> I installed OpenCV 3.4.2 this time around.
- How to Install Caffe and PyCaffe on Jetson TX2
- How to Capture and Display Camera Video with Python on Jetson TX2
- How to Capture Camera Video and Do Caffe Inferencing with Python on Jetson TX2
- Multi-threaded Camera Caffe Inferencing
- Single Shot MultiBox Detector (SSD) on Jetson TX2
- YOLOv3 on Jetson TX2
- How to Do Real-time Object Detection with SSD on Jetson TX2
I also plan to test out NVIDIA’s recently released TensorFlow/TensorRT Models on Jetson on JetPack-3.3, and post an update soon. So stay tuned.