Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
| poc:2025:ondevice-sleepstaging [2025/08/19 10:52] – [Hardware Limitations] fabricio | poc:2025:ondevice-sleepstaging [2025/08/19 13:06] (current) – fabricio | ||
|---|---|---|---|
| Line 5: | Line 5: | ||
| - | * **Status**: | + | * **Status**: |
| * **Duration**: | * **Duration**: | ||
| * **Type**: Hardware/ | * **Type**: Hardware/ | ||
| Line 31: | Line 31: | ||
| * Unclear how the DSP interacts with the NN.. | * Unclear how the DSP interacts with the NN.. | ||
| * Dynamic USleep would be probably out of scope | * Dynamic USleep would be probably out of scope | ||
| - | ==== What Didn't Work ==== | + | |
| - | * Failed approaches (and why) | + | ==== Possible Solution Approaches |
| - | * Performance bottlenecks discovered | + | |
| - | * Incompatibilities found | + | |
| + | | ||
| ==== Surprises & Insights ==== | ==== Surprises & Insights ==== | ||
| - | * Unexpected discoveries | + | * It's barely deep learning anymore. Much more focus on DSP. |
| - | * Hidden complexities | + | * I have absolutely no idea what happens with the performance on such small models. One would need a full development cycle for a new sleep staging model, completely decoupled from recent advances in sleep staging. |
| - | * Simpler alternatives found | + | |
| ===== Recommendations ===== | ===== Recommendations ===== | ||
| - | **Should we proceed? | + | **Should we proceed?**: No, currently out of scope. |
| - | + | ||
| - | **If yes, what needs to happen?** | + | |
| ===== External Resources ===== | ===== External Resources ===== | ||
| - | * [[https://docs.example.com|Official Documentation]] | + | * [[https://www.espressif.com/ |
| - | * [[https://stackoverflow.com/questions/xxx|Relevant SO Discussion]] | + | * [[https://arxiv.org/abs/2212.03332|Edge Impulse MLOps Arxiv Article]] |
| - | * [[Internal: | + | |