Mobile Development with Fexingo: iOS, Android, and App Building Conversations

Episode 79 of Mobile Development with Fexingo dives into how on-device AI is enabling real-time skin cancer screening directly from a smartphone camera. Lucas and Luna explore the technical architecture behind apps like SkinCheck and DermaScan, which use convolutional neural networks (CNNs) trained on dermatoscopic images to classify moles and lesions in under two seconds. They discuss the trade-o...

In this episode of Mobile Development with Fexingo, Lucas and Luna dive into how modern mobile apps are using on-device AI to transcribe music in real time—converting live audio from a guitar, piano, or even a full song into sheet music or MIDI data. They focus on a specific case: the open-source framework Basic Pitch by Spotify, which has been adapted for mobile. Lucas explains how these models r...

In this episode of Mobile Development with Fexingo, Lucas and Luna explore how on-device AI is enabling real-time handwriting recognition in mobile apps. They focus on Apple's PencilKit and the MyScript SDK, discussing the shift from cloud-based OCR to on-device neural networks that process strokes as you write. Lucas explains how transformer models now run on the Neural Engine, achieving latency...

In this episode, Lucas and Luna explore how on-device AI is transforming text-to-speech in mobile apps, analyzing the case of a popular reading app that replaced cloud-based TTS with a neural engine running entirely on the phone. They discuss the latency improvements—from 500 milliseconds down to 50 milliseconds—and the privacy benefits of keeping audio synthesis local. The conversation also cover...

In this episode of Mobile Development with Fexingo, Lucas and Luna dive into how modern mobile apps use on-device AI to analyze sports poses in real time. They break down the technology behind apps like HomeCourt and SwingVision, which track basketball shots and tennis swings without sending video to the cloud. Lucas explains the key role of convolutional neural networks and pose estimation models...

Lucas and Luna explore how on-device machine learning on smartphones now identifies plant diseases in real time. They break down the technical pipeline: how models are trained on thousands of leaf images, compressed via quantization and pruning to run inference in under 200 milliseconds on a phone's neural engine, and integrated into camera apps for instant diagnosis. They discuss real-world deplo...

In Episode 73 of Mobile Development with Fexingo, Lucas and Luna explore how on-device AI enables real-time fall detection in mobile apps. They break down the sensor fusion pipeline — accelerometer, gyroscope, and machine learning models running locally — using Apple's Fall Detection on Apple Watch and Google's Safety Detection API on Android as concrete examples. They discuss the technical challe...

In this episode, Lucas and Luna dive into how on-device AI is enabling real-time sign language translation in mobile apps. They focus on the technical challenges — like hand landmark detection using MediaPipe and transformer-based sequence models that run entirely on the phone without sending data to the cloud. Specific examples include Google's Project Relate and an open-source prototype called S...

In this episode of Mobile Development with Fexingo, Lucas and Luna dive into the controversial and rapidly evolving world of on-device AI voice cloning. They focus on how Apple's Personal Voice feature, announced at WWDC 2023 and now widely available in iOS 18 and 19, allows users to clone their own voice with just 15 minutes of training data — all processed locally on the device. The hosts discus...

In this episode of Mobile Development with Fexingo, Lucas and Luna explore how modern mobile app teams integrate accessibility into UI testing. They dig into real-world tools like Apple's XCUITest with Accessibility Inspector and Google's Espresso with TalkBack, discussing how treating accessibility as a first-class test target catches regressions early and improves UX for everyone. They reference...

In this episode of Mobile Development with Fexingo, Lucas and Luna dive into the world of on-device AI for real-time lip synchronization. They explore how apps like those from Meta and Snapchat use lightweight neural networks to map audio phonemes to mouth shapes without sending data to the cloud. The discussion covers the technical bottlenecks—latency under 200 milliseconds, model size under 10 m...

In episode 68, Lucas and Luna explore how mobile apps are using on-device AI to colorize black-and-white photos in real time. They dive into the technology behind this feature—from convolutional neural networks trained on millions of colorized images to the challenge of maintaining historical accuracy. Lucas shares how Apple and Google have integrated colorization into their native photo editing t...

In this episode of Mobile Development with Fexingo, Lucas and Luna explore how modern mobile apps leverage on-device AI to transform document scanning from a simple photo capture into a powerful tool for text extraction, formatting, and search. They dive into Apple's Vision framework, Google's ML Kit, and real-world apps like Adobe Scan and Microsoft Lens, discussing how these technologies enable...

In this episode of Mobile Development with Fexingo, Lucas and Luna explore how on-device AI powers real-time food recognition in mobile apps. They dive into how Calorie Mama and Snapchat's Dish Scanner use on-device machine learning to identify dishes from a photo, estimate nutritional content, and even suggest recipes. Lucas breaks down the technical stack: lightweight convolutional neural networ...

Episode 65 of Mobile Development with Fexingo dives into on-device AI for real-time 3D body mesh. Lucas and Luna break down how apps like Snapchat and FitOn use neural networks to map thousands of body points without sending data to the cloud. They discuss the tech stack — MediaPipe, ARKit, and TensorFlow Lite — and the engineering trade-offs between accuracy and battery life. Specific focus: how...

In this episode of Mobile Development with Fexingo, Lucas and Luna dive into how modern smartphones leverage on-device AI to optimize camera exposure in real time—no cloud processing required. They break down the mechanics behind Apple's Smart HDR and Google's Real Tone, explaining how neural networks analyze scene luminance, subject motion, and color temperature in milliseconds. The discussion co...

In this episode, Lucas and Luna dive into the on-device AI behind real-time background blur on smartphones. They explore how Apple, Google, and Samsung implement portrait mode for video calls and live streaming, using neural networks and depth sensors. The discussion covers the evolution from dual-camera setups to single-camera solutions with machine learning, the role of the Neural Engine and Ten...

Episode 62 of Mobile Development with Fexingo dives into the nuts and bolts of on-device HDR photo stitching. Lucas and Luna break down how modern smartphones capture three separate exposures in under 100 milliseconds, align them pixel-by-pixel using neural networks, and merge them without sending a single image to the cloud. They walk through the actual computational pipeline — from raw sensor re...

Depth mapping is the invisible layer that powers portrait mode, AR measuring tools, and gesture control on your phone. In this episode, Lucas and Luna unpack how LiDAR and neural networks turn a 2D camera feed into a 3D understanding of the world. They walk through Apple's integration of a dot projector and flood illuminator on iPhones starting in 2017, and how the latest 2026 Android flagships fr...

In episode 60 of Mobile Development with Fexingo, Lucas and Luna explore how on-device AI is enabling real-time emotion recognition in mobile apps. They dive into the technical backbone—facial action coding systems running locally on smartphones—and discuss practical use cases from mental health apps to adaptive gaming. The hosts examine how companies like Affectiva and Apple are approaching this...

Lucas and Luna dive into how modern mobile apps use on-device AI for real-time barcode and QR code scanning, beyond simple price checks. They explore the underlying machine learning models that enable instant recognition of damaged or partially obscured codes, and how apps like those for inventory management and contactless payments leverage dedicated neural processing units (NPUs) on recent iPhon...

In this episode of Mobile Development with Fexingo, Lucas and Luna explore how mobile apps are using on-device AI for real-time skin tone analysis. They dive into the technical challenges of accurate skin tone detection across diverse lighting conditions, the shift away from cloud-based APIs for privacy and latency, and how companies like L'Oréal and Apple are implementing this technology in makeu...

In this episode of Mobile Development with Fexingo, Lucas and Luna explore how on-device AI is enabling real-time audio source separation in mobile apps. They focus on a concrete example: a live music app that uses a small transformer model to isolate vocals, drums, and guitar from a song in under 200 milliseconds on an iPhone processor. The hosts break down how this works with neural spectrogram...

In episode 56 of Mobile Development with Fexingo, Lucas and Luna explore how on-device AI is transforming portrait lighting in mobile apps. They dive into Apple's Neural Engine in the A18 chip, which now computes realistic studio lighting effects—like Rembrandt lighting and butterfly lighting—in real time without uploading a single pixel. Lucas breaks down the technical pipeline: how the on-device...

Lucas and Luna explore how on-device AI is transforming handwriting recognition in mobile apps, moving beyond simple text input to real-time, context-aware transcription. They dive into the technology powering Apple's Scribble, Samsung's S Pen features, and third-party apps like GoodNotes and Nebo, explaining how neural networks process strokes and convert them to digital text without sending data...