Pinned Repositories
ColossalAI
Making large AI models cheaper, faster and more accessible
crack_detection
Concrete Crack Detection for Structural Audit
EDSR-Tensorflow
Tensorflow implementation of Enhanced Deep Residual Networks for Single Image Super-Resolution
face_recognition
The world's simplest facial recognition api for Python and the command line
hdrnet_legacy
An implementation of 'Deep Bilateral Learning for Real-Time Image Enhancements', SIGGRAPH 2017
ImageEnhancer
Image Enhancer to reduce noises, color image, or expand resolutions
keras-retinanet
Keras implementation of RetinaNet object detection.
Label-embeddings-in-image-classification
Convolutional Neural Networks (CNNs) are being widely used for various tasks in Computer Vision. We focus on the task of image classification particularly using CNNs with more focus on the relation or similarity between class labels. The similarity between labels is judged using label word embeddings and incorporated into the loss layer. We propose that shallower networks be learnt with more complex and structured losses, in order to gain from shorter training time and equivalent complexity. We train two variants of CNNs with multiple architectures , all limited to a maximum of ten convolution layers to obtain an accuracy of 93.27% on the Fashion-MNIST dataset and 86.40% on the CIFAR 10 dataset. We further probe the adversarial robustness of the model as well the classspecific behavior by visualizing the class confusion matrix.We also show some preliminary results towards extending a trained variant to zero-shot learning.
Mask_RCNN
Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow
tensorflow_retinanet
RetinaNet with Focal Loss implemented by Tensorflow
dongdongrj's Repositories
dongdongrj/ColossalAI
Making large AI models cheaper, faster and more accessible
dongdongrj/crack_detection
Concrete Crack Detection for Structural Audit
dongdongrj/EDSR-Tensorflow
Tensorflow implementation of Enhanced Deep Residual Networks for Single Image Super-Resolution
dongdongrj/face_recognition
The world's simplest facial recognition api for Python and the command line
dongdongrj/hdrnet_legacy
An implementation of 'Deep Bilateral Learning for Real-Time Image Enhancements', SIGGRAPH 2017
dongdongrj/ImageEnhancer
Image Enhancer to reduce noises, color image, or expand resolutions
dongdongrj/keras-retinanet
Keras implementation of RetinaNet object detection.
dongdongrj/Label-embeddings-in-image-classification
Convolutional Neural Networks (CNNs) are being widely used for various tasks in Computer Vision. We focus on the task of image classification particularly using CNNs with more focus on the relation or similarity between class labels. The similarity between labels is judged using label word embeddings and incorporated into the loss layer. We propose that shallower networks be learnt with more complex and structured losses, in order to gain from shorter training time and equivalent complexity. We train two variants of CNNs with multiple architectures , all limited to a maximum of ten convolution layers to obtain an accuracy of 93.27% on the Fashion-MNIST dataset and 86.40% on the CIFAR 10 dataset. We further probe the adversarial robustness of the model as well the classspecific behavior by visualizing the class confusion matrix.We also show some preliminary results towards extending a trained variant to zero-shot learning.
dongdongrj/Mask_RCNN
Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow
dongdongrj/tensorflow_retinanet
RetinaNet with Focal Loss implemented by Tensorflow
dongdongrj/tf-faster-rcnn
Tensorflow Faster RCNN for Object Detection
dongdongrj/zero-shot-gcn
Zero-Shot Learning with GCN (CVPR 2018)
dongdongrj/ZJL_zero_shot_learning_competition
Code base for ZJL zero shot learning competition.