Merge pull request #167 from arm-university/main

Proj and spotlight updates

Author: matt-cossins

Projects/Projects/Edge-AI-On-Mobile.md

@@ -43,17 +43,20 @@ Other devices also support SME2, including both Apple and Android - [see the ful
 
 ### Project Summary
 
-Select a **mobile edge AI application** that benefits from large matrix operations, multi-modal fusion, or transformer-based processing enabled by SME2. Build and optimize a proof-of-concept application on a vivo X300 phone or other device supporting SME2.
+Select a **mobile edge AI application** that benefits from large matrix operations, multi-modal fusion, or transformer-based processing accelerated by SME2, with real-time speech-to-speech translation, profanity filtering or filler-word removal, and on-device Small Language Models (SLMs) as key example use cases.
 
 Example project areas:  
-    - Real-time video semantic segmentation (e.g., background removal + AR compositing)  
-    - Live object detection + natural-language description (text summary of what the camera sees)  
-    - Multi-sensor fusion (camera + IMU + microphone) for gesture + voice recognition  
-    - On-device lightweight LLM or encoder-only transformer processing for mobile assistants  
+- Real-time speech-to-speech translation
+- On-device SLM for contextual understanding, rewriting, or assistant tasks
+- Profanity detection and filtering (masking, rewriting, audio bleeping)
+- Filler-word removal / speech clean-up (e.g., removing “um”, “uh”, false starts)
+- Real-time video semantic segmentation (e.g., AR captions + background removal)
+- Live object detection with natural-language description
+- Multi-sensor fusion (camera + IMU + microphone) for gesture + voice interaction
 
 Identify a model architecture that maps to wide matrix operations (e.g., ViT, MLP-Mixer, multi-branch CNN with large FC layers). Utilise a mobile-friendly framework (e.g., ExecuTorch, LiteRT, ONNX Runtime, MediaPipe) to leverage SME2 optimizations. Optimize quantization, memory layout, and verify that the large matrix multiplications get scheduled efficiently on the SME2-enabled CPU. Build a mobile app (Android or iOS) that executes the model and utilises it for a compelling use-case.
 
-Utilise the resources and learning paths below and create an exciting and challenging application. Optionally, you could also compare performance vs a reference phone without SME2.
+Utilise the resources and learning paths below and create an exciting and challenging application. As a bonus, you could compare performance vs a reference phone without SME2.
 
 ## Resources from Arm and our partners