nuttx-apps/mlearning
xinhaiteng a37ca36a8a Neon optimized Conv operator
Based on CMSIS-NN, the Conv operator was optimized. Using Neon acceleration, multiply 8 input data and 8 filter data in a single loop; Using Im2col technology, convert the output data into a matrix, calculate 2 rows of input data and 4 rows of filter data in a single large loop, and obtain 2x4 output data.

Signed-off-by: xinhaiteng <xinhaiteng@xiaomi.com>
2024-10-18 09:40:17 +08:00
..
cmsis Kconfig: Make indent consistent with TABS 2023-02-10 14:11:48 +08:00
cmsis-nn Add configurations files. 2024-10-18 09:40:17 +08:00
darknet Kconfig: Make indent consistent with TABS 2023-02-10 14:11:48 +08:00
libnnablart build/Kconfig: fix warnings detected by kconfiglib 2023-02-09 19:45:36 +08:00
tflite-micro Neon optimized Conv operator 2024-10-18 09:40:17 +08:00
.gitignore add Machine Learning folder and menu 2022-01-27 17:33:16 +08:00
CMakeLists.txt add initial cmake build system 2023-07-08 13:52:02 +08:00
Make.defs add Machine Learning folder and menu 2022-01-27 17:33:16 +08:00
Makefile add Machine Learning folder and menu 2022-01-27 17:33:16 +08:00