Our first samples of inference
The first examples of inference from our STM32 N6 board and our analog temperature gauge.
The first examples of inference from our STM32 N6 board and our analog temperature gauge.
The very first image I’ve been able to capture with my B-CAMS-IMX camera on my STM32 N6 board.
Our first ML paper was rejected, but detailed reviewer feedback and an extended deadline gave us a second chance to redesign the model and try again.
First impressions of the STM32 N6 Nucleo board, from the new CubeMX workflow to firmware signing, BOOT1 handling, and ThreadX.
Reflections on finishing CS231n, revisiting The E-Myth, and clarifying a long-term direction in applied AI and engineering work.
A new-year update on getting back into embedded systems and ML foundations after receiving new parts from Mouser.
A year-end update on recovering from illness, finishing the first draft of an embedded AI paper, and looking ahead with optimism.
Fine-tuning a CNN temperature model, dealing with Colab limitations, and noticing how mixed data sources affect prediction accuracy.
A progress update on a temperature-forecasting model that now predicts 15 minutes ahead using a sliding window of past sensor data.
An update on exporting a TensorFlow Lite model into STM32Cube AI and working through code generation and linker issues.