experiencor/image-to-3d-bbox

Stars
237
Rank 169,885 (Top 4 %)
Language
Jupyter Notebook
Created over 7 years ago
Updated over 5 years ago

experiencor/image-to-3d-bbox

experiencor

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Build a CNN network to predict 3D bounding box of car from 2D image.

This is an implementation in Keras of the paper "3D Bounding Box Estimation Using Deep Learning and Geometry" (https://arxiv.org/abs/1612.00496).

Demo on a street video

Usage

"3D Box Regression.ipynb" => to construct and train the network to regress 3D bounding boxes.
"Final KITTI Evaluation.ipynb" => to draw 3D bounding boxes on 2D images.

keras-yolo2

Easy training on custom dataset. Various backends (MobileNet and SqueezeNet) supported. A YOLO demo to detect raccoon run entirely in brower is accessible at https://git.io/vF7vI (not on Windows).

Jupyter Notebook

keras-yolo3

Training and Detecting Objects with YOLO3

self-driving-toy-car

A self driving toy car using end-to-end learning

Jupyter Notebook

deep-viz-keras

Implementations of some popular Saliency Maps in Keras

Jupyter Notebook

semantic-lane-detection

A robust lane detection system based on fully convolutional network for segmenting the road and the lane.

Jupyter Notebook

kangaroo

A dataset of 183 images and annotations for kangaroo detection.

speed-prediction

Use 3D convolutional network to predict the speed of a car via dashcam stream. The final mean squared error on validation set is 3.1 (mile/hour)^2.

Jupyter Notebook

online-machine-learning

Implementation of an online learning algorithm to do classification under concept drift

Jupyter Notebook

basic-rnn-lstm

Implementations of vanilla RNN and LSTM from scratch.

Jupyter Notebook

basic-machine-learning

Implementations of basic machine learning algorithms

Jupyter Notebook

rasa-play

deep-fake