Hyeonki Hong

Results 63 comments of Hyeonki Hong

I'm working on yolov4 v3.0. I will try after release.

Can you share 1 channel dataset?

Training-related functions will not work. I'm very busy these days, so I don't know how long the update will be delayed.

🤩 oh... I got to know this time. I don't have USB-type edgeTPU. It's a feature I want to test. When I'm done with what I'm doing now, I'll buy...

For memory-related issues, I have to test, but I don't have time to spare. Sorry.. :cry:

Not yet. Are the evaluation formulas and dataset for darknet and tflite the same? https://github.com/hunglc007/tensorflow-yolov4-tflite/issues/165

Several tests are in progress on the TPU. I will comment 1~2 days later.

The yolov4-tiny full int8 model is shown below. The red area is the part that cannot be converted. ![yolov4-tiny-relu](https://user-images.githubusercontent.com/21096989/89824531-38065700-db8e-11ea-84bc-2fae33d22125.jpg)

There was a test script, but I forgot to tell it. https://github.com/hhk7734/tensorflow-yolov4/blob/master/test/make_edgetpu_tflite.ipynb

```shell ❯ edgetpu_compiler yolov4-tiny-relu.tflite Edge TPU Compiler version 14.1.317412892 Model compiled successfully in 764 ms. Input model: yolov4-tiny-relu.tflite Input size: 5.96MiB Output model: yolov4-tiny-relu_edgetpu.tflite Output size: 6.19MiB On-chip memory used...