feat: Revamp TFLite & Enable Machine learning Guide for SITARA

[cshilwant]

feat(linux): Revamp TFLite Machine Learning guide

• Revamp TFLite Machine Learning guide which is mainly
  referenced by Sitara platforms.

Signed-off-by: Chirag Shilwant c-shilwant@ti.com

---

feat(configs): Enable Machine learning guide for Sitara

• Enable Foundational_Components_Machine_Learning & all
  the sub-pages in linux toc tree for SITARA platforms.

Signed-off-by: Chirag Shilwant c-shilwant@ti.com

[cshilwant]

Preview available,

AM62X - https://cshilwant.github.io/processor-sdk-doc/processor-sdk-linux-AM62X/esd/docs/11_00_09_04/linux/Foundational_Components/Machine_Learning/tflite.html

AM62PX - https://cshilwant.github.io/processor-sdk-doc/processor-sdk-linux-AM62PX/esd/docs/11_00_09_04/linux/Foundational_Components/Machine_Learning/tflite.html

AM64X - https://cshilwant.github.io/processor-sdk-doc/processor-sdk-linux-AM64X/esd/docs/11_00_09_04/linux/Foundational_Components/Machine_Learning/tflite.html

[praneethbajjuri]

Is any applicable to armv7 TI devices?

[praneethbajjuri]

Co-pilot says the following:

Here are the grammatical issues found in the document:

Line 10: "open source" should be hyphenated as "open-source".
Line 13: The link should be formatted properly (e.g., https://www.tensorflow.org/lite__).
Line 59: "XNNPACK delegate is not supported for ARMv7 based platforms like AM335x & AM437x." should be "The XNNPACK delegate is not supported for ARMv7-based platforms like AM335x and AM437x."
Line 66: "by default contains a pre-installed benchmarking application named as" should be "by default contains a pre-installed benchmarking application named".
Line 110: "This is the path to the benchmark_model binary, which is used to benchmark TensorFlow Lite models." should be "This is the path to the benchmark_model binary, which is used to benchmark TensorFlow Lite models."
Line 112: "Specifies the path to the TFLite model file to be benchmarked." should be "Specifies the path to the TFLite model file to be benchmarked."
Line 115: "The output of the benchmarking application should be similar to," should be "The output of the benchmarking application should be similar to:"
Line 218: "The above GStreamer pipeline reads an H.264 video file, decodes it, and processes it for object detection using a TensorFlow Lite model, displaying bounding boxes around detected objects." should be "The above GStreamer pipeline reads an H.264 video file, decodes it, and processes it for object detection using a TensorFlow Lite model, displaying bounding boxes around detected objects."
These corrections should improve the readability and accuracy of the document.

[cshilwant]

Is any applicable to armv7 TI devices?

@praneethbajjuri yes, all are applicable to armv7 but needs a documentation update with Perf Numbers, etc
Planning to enable these sections once the SDK release for 335x, 437x & 65x is planned

[cshilwant]

@praneethbajjuri Updated the PR with rest all suggestions expect following (as they are exactly the same sentence before vs after)

---

Line 110: "This is the path to the benchmark_model binary, which is used to benchmark TensorFlow Lite models." should be "This is the path to the benchmark_model binary, which is used to benchmark TensorFlow Lite models."

Line 112: "Specifies the path to the TFLite model file to be benchmarked." should be "Specifies the path to the TFLite model file to be benchmarked."

Line 218: "The above GStreamer pipeline reads an H.264 video file, decodes it, and processes it for object detection using a TensorFlow Lite model, displaying bounding boxes around detected objects." should be "The above GStreamer pipeline reads an H.264 video file, decodes it, and processes it for object detection using a TensorFlow Lite model, displaying bounding boxes around detected objects."