User Tools
products:sbc:vim4:npu:demos:yolov7-tiny
Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
products:sbc:vim4:npu:demos:yolov7-tiny [2023/09/18 02:03] sravan [YOLOv7 Tiny VIM4 Demo - 1] |
products:sbc:vim4:npu:demos:yolov7-tiny [2026/04/02 02:46] (current) nick |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| ~~tag> NPU YOLO OpenCV VIM4 ~~ | ~~tag> NPU YOLO OpenCV VIM4 ~~ | ||
| + | |||
| + | **Doc for version ddk-3.4.7.7** | ||
| ====== YOLOv7-tiny VIM4 Demo - 1 ====== | ====== YOLOv7-tiny VIM4 Demo - 1 ====== | ||
| {{indexmenu_n> | {{indexmenu_n> | ||
| + | |||
| + | ===== Introduction ===== | ||
| + | |||
| + | YOLOv7-Tiny is an object detection model. It uses bounding boxes to precisely draw each object in image. | ||
| + | |||
| + | Inference results on VIM4. | ||
| + | |||
| + | {{: | ||
| + | |||
| + | **Inference speed test**: USB camera about **126ms** per frame. | ||
| ===== Train the model ===== | ===== Train the model ===== | ||
| Line 20: | Line 32: | ||
| Follow Docker official documentation to install Docker: [[https:// | Follow Docker official documentation to install Docker: [[https:// | ||
| - | Then fetch the prebuilt NPU Docker | + | Follow |
| ```shell | ```shell | ||
| - | $ docker pull yanwyb/npu:v1 | + | docker pull numbqq/npu-vim4 |
| - | $ docker run -it --name | + | |
| - | -v / | + | |
| - | -v / | + | |
| - | yanwyb/ | + | |
| ``` | ``` | ||
| ==== Get the conversion tool ==== | ==== Get the conversion tool ==== | ||
| - | Download The conversion tool from [[gl>khadas/vim4_npu_sdk]]. | + | You can find the SDK here: [[dl>products/ |
| ```shell | ```shell | ||
| - | $ git clone https://gitlab.com/khadas/ | + | $ wget https://dl.khadas.com/products/ |
| + | $ tar xvzf vim4_npu_sdk-ddk-3.4.7.7-250508.tgz | ||
| + | $ cd vim4_npu_sdk-ddk-3.4.7.7-250508 | ||
| + | $ ls | ||
| + | adla-toolkit-binary | ||
| ``` | ``` | ||
| + | |||
| + | * '' | ||
| + | * '' | ||
| + | * '' | ||
| + | * '' | ||
| + | |||
| + | <WRAP important> | ||
| + | If your kernel is older than 241129, please use branch npu-ddk-1.7.5.5. | ||
| + | </ | ||
| ==== Convert ==== | ==== Convert ==== | ||
| Line 56: | Line 77: | ||
| if self.grid[i].shape[2: | if self.grid[i].shape[2: | ||
| ``` | ``` | ||
| + | |||
| + | <WRAP important> | ||
| + | yolo.py has many forward. Right place is class **IDetect** function **fuseforward**. | ||
| + | </ | ||
| Then, run '' | Then, run '' | ||
| Line 84: | Line 109: | ||
| --quantize-dtype int8 --outdir onnx_output | --quantize-dtype int8 --outdir onnx_output | ||
| --channel-mean-value " | --channel-mean-value " | ||
| + | --inference-input-type " | ||
| + | --inference-output-type " | ||
| --source-file dataset.txt | --source-file dataset.txt | ||
| --batch-size 1 --target-platform PRODUCT_PID0XA003 | --batch-size 1 --target-platform PRODUCT_PID0XA003 | ||
| Line 104: | Line 131: | ||
| ``` | ``` | ||
| + | |||
| + | <WRAP important> | ||
| + | If your kernel is older than 241129, please use version before tag ddk-3.4.7.7. | ||
| + | </ | ||
| ==== Install dependencies ==== | ==== Install dependencies ==== | ||
| Line 127: | Line 158: | ||
| # Run | # Run | ||
| - | $ sudo ./ | + | $ ./ |
| ``` | ``` | ||
| Line 143: | Line 174: | ||
| # Run | # Run | ||
| - | $ sudo ./ | + | $ ./ |
| ``` | ``` | ||
Last modified: 2023/09/18 02:03 by sravan