Build intelligence into your apps using machine learning models from the research community designed for Core ML.
Models are in Core ML format and can be integrated into Xcode projects. You can select different versions of models to optimize for sizes and architectures.
Images
Images
FCRN-DepthPredictionDepth Estimation
Predict the depth from a single image.
MNISTDrawing Classification
Classify a single handwritten digit (supports digits 0-9).
UpdatableDrawingClassifierDrawing Classification
Drawing classifier that learns to recognize new drawings based on a K-Nearest Neighbors model (KNN).
MobileNetV2Image Classification
The MobileNetv2 architecture trained to classify the dominant object in a camera frame or image.
Resnet50Image Classification
A Residual Neural Network that will classify the dominant object in a camera frame or image.
SqueezeNetImage Classification
A small Deep Neural Network architecture that classifies the dominant object in a camera frame or image.
DeeplabV3Image Segmentation
Segment the pixels of a camera frame or image into a predefined set of classes.
YOLOv3Object Detection
Locate and classify 80 different types of objects present in a camera frame or image.
YOLOv3-TinyObject Detection
Locate and classify 80 different types of objects present in a camera frame or image.
PoseNetPose Estimation
Estimates up to 17 joint positions for each person in an image.
Text
Text
BERT-SQuADQuestion Answering
Find answers to questions about paragraphs of text.
FCRN-DepthPrediction
FCRN.mlmodelStoring model weights using full precision (32 bit) floating point numbers. 254.7MB
PoseNetMobileNet075S8FP16.mlmodelThis model uses a MobileNetV1 architecture with a width multiplier of 0.75 and an output stride of 8, storing its weights using half-precision (16 bit) floating point numbers. 2.6MB
PoseNetMobileNet075S16FP16.mlmodelThis model uses a MobileNetV1 architecture with a width multiplier of 0.75 and an output stride of 16, storing its weights using half-precision (16 bit) floating point numbers. 2.6MB
PoseNetMobileNet100S8FP16.mlmodelThis model uses a MobileNetV1 architecture with a width multiplier of 1.00 and an output stride of 8, storing its weights using half-precision (16 bit) floating point numbers. 6.7MB
PoseNetMobileNet100S16FP16.mlmodelThis model uses a MobileNetV1 architecture with a width multiplier of 1.00 and an output stride of 16, storing its weights using half-precision (16 bit) floating point numbers. 6.7MB
Finding Answers to Questions in a Text DocumentLocate relevant passages in a document by asking the Bidirectional Encoder Representations from Transformers (BERT) model a question.