ユーザーにCreateMLを使わせてカスタマイズを実現するワークフローを考えている

Wow, well designed UI, are you trying to revive the CreateML App?

Just how we have CreateML for CoreML, we desperately need a CreateAI for CoreAI - would make life so easy!

The Reality Composer app has an 3D Object Capture mode with a bounding box! I believe you can use that as input for the CreateML

Create ML gets major updates for image input workflows. The WWDC session on machine learning tools showed several improvements to Create ML that make it more practical for indie developers and small teams. What's new: → Drag-and-drop image folders for training (no more manual labeling for classification) → Auto-labeling suggestions based on folder structure → Image augmentation built-in (rotation, lighting, perspective) → Direct export to Core ML format optimized for Neural Engine → Integration with RealityKit for AR use cases The practical impact: you can now train a custom image classifier in hours, not days. The workflow is: 1. Drop 50 images per category into folders 2. Create ML infers labels from folder names 3. Train (runs locally on your Mac) 4. Export to Core ML 5. Drop into your app For indie game devs: this is the workflow for training your own sprite/asset classifiers. You can build a "find similar art" feature into your asset library without writing ML code. For AR apps: Create ML can now train models that recognize 3D objects, not just 2D images. You can build an AR experience that recognizes your specific product line. The Neural Engine optimization is the underrated part. Models trained with Create ML now run 3-4x faster on iPhone because they're compiled for the dedicated ML hardware. The battery impact is also much lower. For accessibility apps: this is huge. You can now build a custom object recognizer for a specific use case (recognizing medication, identifying tools for someone with vision impairment, etc.) without an ML PhD. The session is in the WWDC 2026 video archive. Worth watching for the full workflow demo. #WWDC26 #CreateML #MachineLearning #Developer

boredom is scarce in a age where quick ideas are easily built with ai (i.e. the last few hours I built a supplement to createML on mac.. something that I would of thought of but never done) i miss when I was a sophomore in high school having to project out ideas/scenarios

Apple provides MLSoundClassifer in the CreateML framework. It’s a model you train with audio files to recognize and identify sounds in your environment and it’s on-device eg its used in iOS for recognition of alarms, animals and common household sounds developer.apple.com/document…

It’s the weekend, exploring new app idea. So we train a custom Golf AI model with CreateML. ⛳️

RectLabel is the app you need for any image classification, detection, or custom ML training. So powerful and useful for small experiments. RectLabel helped me a ton when I created a spider detector with CreateML. Works in realtime. Amazing what CoreML enables.

createMLを初めて使った。画像解析やってみたかったから、Mac1台でできるの凄すぎる、！codeXも初めて使ったけど、えぐいなあれ

createMLは小さいもの相手だと想像したような追従性が出にくいなぁ

I've been building panlabel — a fast Rust CLI that converts between dataset annotation formats — and I'm a few releases behind on sharing updates. Here's a quick catch-up. v0.3.0 added Hugging Face ImageFolder support, including remote Hub import via --hf-repo. You can point it at a HF dataset repo and it figures out the layout (metadata.jsonl, parquet shards, even zip-style splits that contain YOLO or COCO inside). v0.4.0 overhauled auto-detection so it gives you concrete evidence when format detection is ambiguous ("found YOLO labels/ but missing images/") instead of a generic error. Also added Docker images. v0.5.0 brought split-aware YOLO reading for Roboflow/Ultralytics Hub exports and conversion report explainability — every adapter now explains its deterministic policies so you know exactly what happens to your data. v0.6.0 is the big one. Five new format adapters: → LabelMe JSON (per-image, with polygon-to-bbox envelope) → Apple CreateML JSON (center-based coords) → KITTI (autonomous driving standard — 15 fields per line) → VGG Image Annotator (VIA) JSON → RetinaNet Keras CSV That brings panlabel to 13 supported formats with full read, write, and auto-detection. Also in v0.6.0: YOLO confidence token support, dry-run mode for previewing conversions, and content-based CSV detection. Single binary, no Python dependencies. Install via pip, brew, cargo, or grab a pre-built binary from GitHub releases. This is the kind of project I enjoy just steadily plodding away at — ticking off one format at a time until every common object detection annotation format is covered. Still sticking with detection bboxes for now, but the format list keeps growing. #ObjectDetection #Rust #MachineLearning #ComputerVision #OpenSource

・なぜUnityなのか？ ・Galaxy XRを韓国まで買付に行ったとのこと（今日はないです） ・CreateML、Scaniverseのモデルを位置合わせして表示、ボンネットを閉じると透けているような表示に

おはようございます☀ CreateMLに関して登壇しようとおもっていますが、まだ検証が終わっていない...！ 早くやります

Updated for iOS 26! ⚡️ Learn how to train and integrate Custom ML Models with Create ML, all on-device:  kodeco.com/ios/paths/apple-a… #iOS26 #iOSDev #SwiftLang #AppleAI #CreateML #OnDeviceAI #MachineLearning #Kodeco #LearnToCode

次は @devjonny_ さんのMachine Learning with CreateML #extension_dc

這兩天的下班時間，試著用 CreateML 拿來跑內部小專案的數據預測，意外有趣，想不到第一次 vibe coding 是用在 ML 上面 😆

Two words... Holy Sh*t. Less than 30 days in and already learning about Machine Learning via CreateML... completely bonkers 😂 I FEEL UNSTOPPABLE!!! MWAhaha. Date types r cool too... 🎉 I just finished Day 26 of the #100DaysOfSwiftUI at hackingwithswift.com/100/swi… via @twostraws

Apple Vision Proで実車のシースルー表示を試してみた。ボンネットを閉めても内部が見える。ObjectTrackingを利用して位置合わせ トラッキング用3Dデータの作成は #scaniverse を利用 #visionpro #createml #MINI