Repository navigation
Conversation
Authored with assistance from OpenAI Codex.
Document the tested camera, NPU and UART path with verified programming and reference checks, and retain the optional display and strict PT2E comparison limitations. Authored with assistance from OpenAI Codex.
usamahz
marked this pull request as ready for review
November 18, 2026 14:50
Align the MobileSAM Learning Path metadata with the upstream category validator. Authored with assistance from OpenAI Codex.
usamahz
force-pushed
the
codex/mobilesam-alif-learning-path
branch
from
November 18, 2026 15:34
eca2b5f to
46787d9
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds a Learning Path for deploying MobileSAM on the Alif Ensemble E8 A0 HP core and Ethos-U85. The supported default captures camera frames, runs the four-prompt model, and reports results through UART. Review the lessons in navigation order, followed by the downloadable application/vendor patches and MRAM programming helper. The workflow uses released dependencies, a pinned public calibration image, firmware memory checks, and SWD/OSPI programming with independent readbacks.
A fresh export passed the unchanged public-image quality gate, followed by tensor-contract checks and a clean firmware build. The default camera/NPU/UART firmware completed 49 frames across two reset-pin runs with no reported errors, then seven more frames after debugger capture and resume. Captured frame 23 matched the retained integer TOSA reference byte-for-byte for all mask and quality outputs; the full programmed model and firmware were independently read back. These are bounded observations, not a long-term reliability or scene-accuracy claim. The nine-page Hugo preview, navigation, links, and downloadable asset hashes passed validation.
The earlier strict PT2E comparison failed and remains reported unchanged; the separate integer TOSA result does not turn that test into a pass. The live camera dashboard was visible and updating in earlier E8 runs, while DSI ACK/packet errors appeared in the logs. The rebuilt optional dashboard configuration has not been physically validated, and the supported default disables the display.
Authored with assistance from OpenAI Codex.
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of the Creative Commons Attribution 4.0 International License.