Object recognition
Library of object recognition models for object detection and instance segmentation. It includes the original paper, pretrained weights available, and deep learning framework used by code shared.
===
- PT = PyTorch
- TF = TensorFlow
- C = Caffe
- MXN = MXNet
- MCN = MaxConvNet
- T = Torch
- ~ Code not very accurate / incomplete
- / Not available
| No | Name | Year | Paper | PT | TF | C | MXN | MCN | T | Pretrained weights |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Mask R-CNN | 2017 | * | PT | TF | C | PyTorch, Tensorflow | |||
| 2 | FCIS: Fully convolutional instance-aware semantic segmentation | 2016 | * | MXN | MXNet | |||||
| 3 | Instance-sensitive Fully Convolutional Networks | 2016 | * | / | ||||||
| 4 | Object detection via region-based fully convolutional networks | 2016 | * | C | Caffe | |||||
| 5 | MNC: Instance-aware semantic segmentation via multi-task network cascades | 2015 | * | C | Caffe | |||||
| 6 | R-FCN: Object detection via region-based fully convolutional networks | 2016 | * | TF | C | MXN | PyTorch, TensorFlow, Caffe, MxNet | |||
| 7 | Instancecut: from edges to instances with multicut | 2016 | * | / | ||||||
| 8 | Deep Watershed Transform for Instance Segmentation | 2016 | * | TF | TensorFlow | |||||
| 9 | Pixelwise Instance Segmentation with a Dynamically Instantiated Network | 2017 | * | / | ||||||
| 10 | SGN: Sequential Grouping Networks for Instance Segmentation | 2017 | * | / | ||||||
| 11 | DeepMask: Learning to segment object candidates | 2015 | * | PT | T | PyTorch, Torch | ||||
| 12 | SharpMask: Learning to refine object segments | 2016 | * | TF | T | TensorFlow , Torch | ||||
| 13 | A MultiPath Network for Object Detection | 2016 | * | T | Torch | |||||
| 14 | Iterative instance segmentation | 2015 | * | |||||||
| 15 | Recurrent Instance Segmentation | 2015 | * | T | / | |||||
| 16 | MaskLab: Instance Segmentation by Refining Object Detection with Semantic and Direction Features | 2017 | * | / | ||||||
| 17 | Semantic Instance Segmentation via Deep Metric Learning | 2017 | * | ~PT | / | |||||
| 18 | Recurrent Pixel Embedding for Instance Grouping | 2017 | * | MCN | MatConvNet | |||||
| 19 | End-to-End Instance Segmentation with Recurrent Attention | 2017 | * | TF | TensorFlow | |||||
| 20 | PANet: Path Aggregation Network for Instance Segmentation | 2018 | * | PT | PyTorch | |||||
| 21 | BlitzNet: A Real-Time Deep Network for Scene Understanding | 2017 | * | TF | TensorFlow | |||||
| 22 | FastMask: Segment Multi-scale Object Candidates in One Shot | 2017 | * | C | Caffe | |||||
| 23 | Deep Residual Learning for Image Recognition | 2016 | * | C | ||||||
| 24 | Feature pyramid networks for object detection | 2016 | * | / | ||||||
| 25 | Deep Residual Learning for Image Recognition | 2015 | * | TF | MCN | T | TensorFlow, MatConvNet, Torch | |||
| 26 | Cascade R-CNN: Delving into High Quality Object Detection | 2017 | * | C | Caffe | |||||
| 27 | FastMask: Segment Multi-scale Object Candidates in One Shot | 2016 | * | C | Caffe | |||||
| 28 | Pseudo Mask Augmented Object Detection | 2018 | * | / | ||||||
| 29 | One-Shot Instance Segmentation | 2018 | * | TF | TensorFlow | |||||
| 30 | YOLACT: Real-time Instance Segmentation | 2019 | * | PT | PyTorch | |||||
| 31 | Instance-Level Salient Object Segmentation | 2017 | * | / | ||||||
| 32 | Semantic Instance Segmentation with a Discriminative Loss Function | 2017 | * | |||||||
| 33 | Recurrent Neural Networks for Semantic Instance Segmentation | 2017 | * | PT | PyTorch | |||||
| 34 | One-Shot Instance Segmentation (Siamese Mask R-CNN) | 2018 | * | TF | TensorFlow | |||||
| 35 | Affinity Derivation and Graph Merge forInstance Segmentation | 2018 | * | TF | TensorFlow | |||||
| 36 | DASNet: Reducing Pixel-level Annotations forInstance and Semantic Segmentation | 2018 | * | / | ||||||
| 37 | Predicting Future Instance Segmentation by Forecasting Convolutional Feature | 2018 | * | C | Caffe | |||||
| 38 | RetinaMask: Learning to predict masks improves state-of-the-art single-shot detection for free | 2019 | PT | PyTorch | ||||||
| A1 | Convolutional feature masking for joint object and stuff segmentation | 2014 | * | |||||||
| A2 | Rich feature hierarchies for accurate object detection and semantic segmentation | 2013 | * | ~PT | / | |||||
| A3 | Simultaneous detection and segmentation | 2014 | * | |||||||
| A4 | Hypercolumns for object segmentation and fine-grained localization | 2014 | * |