IceVision: Agnostic Object Detection Framework
- End-to-End Object Detection Framework - Supports Fastai and Lightning with more to come - Unified Data API (Annotation Parsers) - Datasets Hub - ...
object-detection computer-vision fastai pytorch-lightning coco-datasets voc-datasets coco-parser voc-parser custom-parser efficientdet faster-rcnn mask-rcnn annotation-parsers article code notebook demo done wandb library research

Why IceVision ?

  • IceVision is an Object-Detection Framework that connects to different libraries/frameworks such as Fastai, Pytorch Lightning, and Pytorch with more to come.
  • Features a Unified Data API with out-of-the-box support for common annotation formats (COCO, VOC, etc.)
  • The IceData repo hosts community maintained parsers and custom datasets
  • Provides flexible model implementations with pluggable backbones
  • Helps researchers reproduce, replicate, and go beyond published models
  • Enables practioners to get moving with object detection technology quickly

Models supported currently

  • EfficientDet - Ross Wightman's Pytorch Implementation
  • Faster RCNN with FPN backbones.
  • Mask RCNN with FPN Backbones.

Toy Datasets provided

  • COCO
  • PennFudan
  • PETS
  • Pascal VOC 2012
  • Object Fridge
  • Birds


Don't forget to tag @oke-aditya , @lgvaz , @ai-fast-track , @paras-jain in your comment, otherwise they may not be notified.

Authors original post
Deep Learning | Computer Vision
Deep Learning practitioner.
Helping democratize AI/DL. Breaking down complex concepts into easy to understand pieces. PhD - Biomed. Eng. Postdoc - Biomed. Eng. & Neurosciences Created StatMap3D - DataFinder
Interested in Vision, Natural Language Processing and Social network dynamics.
Share this project
Similar projects
Different deep learning architectures definitions that can be applied to image segmentation.
Training a pets detector model with TFOD API (TF 2)
In this notebook, we will be training a custom object detection model using the latest TensorFlow Object Detection (TFOD) API which is based on TensorFlow ...
Image and Bounding Box Annotation Slicer
This easy-to-use library slices (also resizes) images and its bounding box annotations into tiles of specific sizes or any arbitrary number of equal parts. ...
Object Detection for Dummies
We will go through several basic concepts, algorithms, and popular deep learning models for image processing and object detection.
Top collections