Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Full YOLOv3 Support

See original GitHub issue

Start with the `why`:

Although DepthAI already supports tiny-YOLOv3 (implemented in https://github.com/luxonis/depthai/pull/131), it can be useful in some cases it is advantages to use full YOLOv3 instead. Example include as when higher accuracy, more classes, or smaller objects need to be identified take precedent over the faster framerate of tinyYOLOv3.

Move to the `how`:

We already support the model itself, so likely the only additional items needed is the tensor output processing on the host.

Ideally, we should do this on the host to start, and then later support it on the DepthAI standalone as we are doing for tinyYOLOv3 in https://github.com/luxonis/depthai/issues/137.

Move to the `what`:

Implement support for YOLOv3 on DepthAI.

Issue Analytics

State:
Created 3 years ago
Comments:7

Top GitHub Comments

1reaction

szabi-luxoniscommented, Oct 6, 2020

Support added in #202

1reaction

zhijietan94commented, Sep 23, 2020

Looks great! Would like to try it but the google drive link for YOLOv3 blob is restricted access. @Luxonis-Brandon Can you help to adjust the access rights for download? Thanks!

Top Results From Across the Web

YOLOv3: Real-Time Object Detection Algorithm (Guide) - viso.ai

It allows the model to look at the whole image at test time, so its predictions are informed by the global context in...

YOLO: Real-Time Object Detection - Joseph Redmon

YOLOv3 uses a few tricks to improve training and increase performance, including: multi-scale predictions, a better backbone classifier, and more. The full ......

What's new in YOLO v3? - Towards Data Science

A review of the YOLO v3 object detection algorithm, covering new features, performance benchmarks, and link to the code in PyTorch.

[1804.02767] YOLOv3: An Incremental Improvement - arXiv

At 320x320 YOLOv3 runs in 22 ms at 28.2 mAP, as accurate as SSD but three times faster. When we look at the...

How to Perform Object Detection With YOLOv3 in Keras

Since the whole detection pipeline is a single network, ... have used this code as a starting point and updated it to support...