Skip to content

Instantly share code, notes, and snippets.

@tzmartin

tzmartin/roboflow-vision-models.md

Created December 6, 2024 18:42
Download ZIP

Revisions

  1. tzmartin created this gist Dec 6, 2024.
    164 changes: 164 additions & 0 deletions roboflow-vision-models.md
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,164 @@
    # Roboflow Model Library

    The Roboflow Model Library contains pre-configured model architectures for easily training computer vision models.

    ## PyTorch Object Detection

    ### YOLOv8
    - **Description**: YOLOv8 is a state-of-the-art object detection and image segmentation model created by Ultralytics.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/how-to-train-yolov8-on-a-custom-dataset/)
    - [Video](https://youtu.be/wuZtUMEiKWY)
    - [Repo](https://github.com/ultralytics/ultralytics)
    - [Jupyter Notebook](https://colab.research.google.com/github/roboflow-ai/notebooks/blob/main/notebooks/train-yolov8-object-detection-on-custom-dataset.ipynb)

    ### YOLOv5
    - **Description**: A very fast and easy to use PyTorch model that achieves state of the art (or near state of the art) results.
    - **Links**:
    - [Tutorial](https://blog.roboflow.ai/how-to-train-yolov5-on-a-custom-dataset/)
    - [Video](https://youtu.be/MdF6x6ZmLAY)
    - [Repo](https://github.com/ultralytics/yolov5)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1gDZ2xcTOgR39tGGs-EZ6i3RTs16wmzZQ)

    ### YOLOv7
    - **Description**: The latest in the YOLO mainline, from the creators of YOLOv4, YOLOv7 achieves state of the art performance on MS COCO amongst realtime object detectors.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/yolov7-custom-dataset-training-tutorial/)
    - [Video](https://www.youtube.com/watch?v=5nsmXLyDaU4)
    - [Repo](https://github.com/WongKinYiu/yolov7)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1X9A8odmK4k6l26NDviiT6dd6TgR-piOa)

    ### MT-YOLOv6
    - **Description**: MT-YOLOv6, or YOLOv6, is a high performance model in the YOLO family of models. Released in June 2022, it sets a new state of the art.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/how-to-train-yolov6-on-a-custom-dataset/)
    - [Video](https://www.youtube.com/watch?v=fFCWrMFH2UY)
    - [Repo](https://github.com/meituan/YOLOv6)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1YnbqOinBZV-c9I7fk_UL6acgnnmkXDMM)

    ### YOLOv7 leveraging OpenVINO™ Integration with Torch-ORT
    - **Description**: Accelerated Inference of a YOLOv7 Model using OpenVINO™ Integration with Torch-ORT.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/accelerate-pytorch-openvino-torch-ort/)
    - [Repo](https://github.com/roboflow-ai/yolov7)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/gist/mmcquade11/06e97618cdb4739b43807d2b8d655de1/how-to-run-inference-on-yolov7-with-native-pytorch-and-openvino-torch-ort.ipynb)

    ### Scaled-YOLOv4
    - **Description**: As of December 2020, Scaled-YOLOv4 is state-of-the art for object detection. Scaled-YOLOv4 implements YOLOv4 in the PyTorch framework with Cross Stage Partial network layers.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/how-to-train-scaled-yolov4/)
    - [Video](https://www.youtube.com/watch?v=rEbpKxZbvIo)
    - [Repo](https://github.com/WongKinYiu/ScaledYOLOv4)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1LDmg0JRiC2N7_tx8wQoBzTB0jUZhywQr#scrollTo=odKEqYtTgbRc)

    ### YOLOS
    - **Description**: YOLOS is a new transformer based object detection model.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/train-yolos-transformer-custom-dataset)
    - [Video](https://www.youtube.com/watch?v=N0V0xxSi6Xc&t=342s)
    - [Repo](https://github.com/huggingface/transformers)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1kfU7TpWYrZKRR8eGwdvKgpoEPUtC-J6I)

    ### YOLOR
    - **Description**: You Only Learn One Representation (YOLOR) is a state-of-the-art object detection model that pre-trains an implicit knowledge network and a set of parameters to represent explicit knowledge.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/train-yolor-on-a-custom-dataset/)
    - [Video](https://www.youtube.com/watch?v=sZ5DiXDOHEM)
    - [Repo](https://github.com/roboflow-ai/yolor)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1e1Uk9SjxBaagu7aoGZ4oTcqePhnMLM23?usp=sharing)

    ### YOLOX
    - **Description**: YOLOX is the winner of the most recent CMU Streaming Perception Challenge for its ability to tradeoff both edge inference speed and accuracy.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/how-to-train-yolox-on-a-custom-dataset/)
    - [Video](https://www.youtube.com/watch?v=q3RbFbaQQGw&t=15s)
    - [Repo](https://github.com/Megvii-BaseDetection/YOLOX)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1_xkARB35307P0-BTnqMy0flmYrfoYi5T#scrollTo=igwruhYxE_a7)

    ### EfficientDet-D0-D7
    - **Description**: A scalable, state of the art object detection model, implemented here within the TensorFlow 2 Object Detection API.
    - **Links**:
    - [Tutorial](https://blog.roboflow.ai/train-a-tensorflow2-object-detection-model/)
    - [Repo](https://github.com/tensorflow/models/tree/master/research/object_detection)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1sLqFKVV94wm-lglFq_0kGo2ciM0kecWD#scrollTo=fF8ysCfYKgTP&uniqifier=1)

    ### YOLOv4-tiny
    - **Description**: The tiny and fast version of YOLOv4 - good for training and deployment on limited compute resources, and getting a feel for your dataset.
    - **Links**:
    - [Tutorial](https://blog.roboflow.ai/train-yolov4-tiny-on-custom-data-lighting-fast-detection/)
    - [Video](https://www.youtube.com/watch?v=NTnZgLsk_DA&t=0s)
    - [Repo](https://github.com/roboflow-ai/darknet)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1PWOwg038EOGNddf6SXDG5AsC8PIcAe-G#scrollTo=Cdj4tmT5Cmdl)

    ### YOLOv4 Darknet
    - **Description**: YOLOv4 has emerged as the best real time object detection model. YOLOv4 carries forward many of the research contributions of the YOLO family of models along with new modeling and data augmentation techniques. This implementation is in Darknet.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/training-yolov4-on-a-custom-dataset/)
    - [Video](https://www.youtube.com/watch?v=N-GS8cmDPog)
    - [Repo](https://github.com/roboflow-ai/darknet)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1mzL6WyY9BRx4xX476eQdhKDnd_eixBlG#scrollTo=GNVU7eu9CQj3)

    ### YOLOv5-OBB
    - **Description**: Oriented bounding boxes are bounding boxes rotated to better fit the objects represented on an angle.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/yolov5-for-oriented-object-detection/)
    - [Video](https://www.youtube.com/watch?v=iRkCNo9-slY)
    - [Repo](https://github.com/hukaixuan19970627/yolov5_obb)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/16nRwsioEYqWFLBF5VpT_NvELeOeupURM#scrollTo=1NZxhXTMWvek)

    ### Detectron2
    - **Description**: Detectron2 is a model zoo of its own for computer vision models written in PyTorch. Detectron2 includes all the models that were available in the original Detectron, such as Faster R-CNN, Mask R-CNN, RetinaNet, and DensePose. It also features several new models, including Cascade R-CNN, Panoptic FPN, and TensorMask.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/how-to-train-detectron2/)
    - [Video](https://www.youtube.com/watch?v=e8LPflX0nwQ)
    - [Repo](https://github.com/facebookresearch/detectron2)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1-TNOcPm3Jr3fOJG8rnGT9gh60mHUsvaW#scrollTo=kc8MmgZugZWR)

    ### EfficientDet
    - **Description**: EfficientDet achieves the best performance in the fewest training epochs among object detection model architectures, making it a highly scalable architecture especially when operating with limited compute.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/training-efficientdet-object-detection-model-with-a-custom-dataset/)
    - [Video](https://www.youtube.com/watch?v=yJg1FX2goCo)
    - [Repo](https://github.com/roboflow-ai/Monk_Object_Detection)
    - [Jupyter Notebook](https://github.com/roboflow-ai/Monk_Object_Detection/blob/master/Tutorial.ipynb)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1ZmbeTro4SqT7h_TfW63MLdqbrCUk_1br#scrollTo=KwDS9qqBbMQa)

    ### YOLOv4 PyTorch
    - **Description**: YOLOv4 has emerged as one of the best real-time object detection models. YOLOv4 carries forward many of the research contributions of the YOLO family of models along with new modeling and data augmentation techniques. This implementation is in PyTorch.
    - **Links**:
    - [Video](https://www.youtube.com/watch?v=9hVgyeI4g4o)
    - [Repo](https://github.com/roboflow-ai/pytorch-YOLOv4)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1b08y_nUYv5UtDY211NFfINY7Hy_pgZDt)

    ### Faster R-CNN
    - **Description**: One of the most accurate object detection algorithms but requires a lot of power at inference time. A good choice if you can do processing asynchronously on a server.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/training-a-tensorflow-faster-r-cnn-object-detection-model-on-your-own-dataset/)
    - [Video](https://www.youtube.com/watch?v=4OXntFVfFio)
    - [Repo](https://github.com/roboflow-ai/tensorflow-object-detection-faster-rcnn)
    - [Jupyter Notebook](https://github.com/roboflow-ai/tensorflow-object-detection-faster-rcnn/blob/master/Tutorial_Faster_RCNN.ipynb)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1U3fkRu6-hwjk7wWIpg-iylL2u5T9t7rr#scrollTo=yhzxsJb3dpWq)

    ### YOLO v3 PyTorch
    - **Description**: Though it is no longer the most accurate object detection algorithm, YOLO v3 is still a very good choice when you need real-time detection while maintaining excellent accuracy. PyTorch version.
    - **Links**:
    - [Video](https://www.youtube.com/watch?v=lJT_KRFLLRY&feature=youtu.be)
    - [Repo](https://github.com/roboflow-ai/yolov3)
    - [Jupyter Notebook](https://github.com/roboflow-ai/yolov3/blob/master/Tutorial.ipynb)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1ntAL_zI68xfvZ4uCSAF6XT27g0U4mZbW#scrollTo=VHS_o3KGIyXm)

    ### YOLO v3 Keras
    - **Description**: Though it is no longer the most accurate object detection algorithm, YOLO v3 is still a very good choice when you need real-time detection while maintaining excellent accuracy. Keras implementation.
    - **Links**:
    - [Tutorial](https://blog.roboflow.ai/training-a-yolov3-object-detection-model-with-a-custom-dataset/)
    - [Repo](https://github.com/roboflow-ai/keras-yolo3)
    - [Jupyter Notebook](https://github.com/roboflow-ai/keras-yolo3/blob/master/Tutorial.ipynb)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1ByRi9d6_Yzu0nrEKArmLMLuMaZjYfygO#scrollTo=WgHANbxqWJPa)

    ### MobileNetSSDv2
    - **Description**: This architecture provides good realtime results on limited compute. It's designed to run in realtime (30 frames per second) even on mobile devices.
    - **Links**:
    - [Tutorial](https://blog.roboflow.com/training-a-tensorflow-object-detection-model-with-a-custom-dataset/)
    - [Repo](https://github.com/roboflow-ai/tensorflow-object-detection-faster-rcnn)
    - [Jupyter Notebook](https://github.com/roboflow-ai/tensorflow-object-detection-faster-rcnn/blob/master/Tutorial_Mobilenet.ipynb)
    - [Colab Notebook (Free GPU)](https://colab.research.google.com/drive/1wTMIrJhYsQdq_u7ROOkf0Lu_fsX5Mu8a)