Maria Chen

Hi Oded, No worries, there's no need to convert anything. You can use 24bit model as well. When it is being converted using the toolchain, it will automatically fit the NPU format (RGBA8888). You will need to input the right parameters for pre-proc…

in RGB 24-bit support Comment by Maria Chen January 2022

Hi Kendrick, 因為SDK架構的更動，95MB為舊數據，經過一些驗證，新架構的模型上限將會下調 (建議不超過75MB)。至於讓模型變小的事情，建議直接在網路上找比較小的openpose的model，也可以參考這個網站來做pruning: https://github.com/ChrisAllenMing/pruned_lightw…

in Got Error When Trying to load openpose model into KL720 Comment by Maria Chen December 2021

Hi Albert, Unfortunately, the current Kneron_DFUT (v1.2.0) only supports Raspberry Pi 4, Windows, and Ubuntu. There will be a new version of Kneron PLUS, estimated to be released at the end of the month. In that version, there will be a file you co…

in How to setup inference environment on the Raspberry pi 3 with Mini-AI-520 Comment by Maria Chen December 2021

Hi Bo-Chen, 有，可以把host_lib的firmware放在Kneron_DFUT裡適當的路徑下，再用以下的command: $ sudo ./KneronDFUT --kl520-update --port {port ID} --scpu {SCPU_FILE_PATH} --ncpu {NCPU_FILE_PATH} 這個沒有在文檔裡面，不過可以用來將KDP2 firmware變回KDP firm…

in KL520 firmware downgrade Comment by Maria Chen December 2021

Hi Oded, Have you updated firmware to your flash? Have you modified your scpu file, and if so, could you provide it? What example were you running when you ran into the error? We could help you debug if you could provide the firmware and PLUS exampl…

in Error 100 with a customized example Comment by Maria Chen December 2021

Hi Kendrick, 這個error的原因有可能是因為.nef檔太大。雖然沒有超過上限，但是firmware的更動有時會影響model能用的空間大小，所以可以請您試著讓.nef檔變小一點，在測試看看嗎? 或是能不能請您提供.nef模型和KL720_Openpose_inference.py，我們可以試著調整設定然後…

in Got Error When Trying to load openpose model into KL720 Comment by Maria Chen December 2021

還有，請問一下有用 sudo ./scan_devices執行嗎? 另外，可以試著以管理者權限創建 (或開啟)檔案 /etc/udev/rules.d/10-local.rules (若不是用管理者權限開啟的話無法存檔)，在檔案中加入下面三行: SUBSYSTEM=="usb",ATTRS{idVendor}=="3231",ATTRS…