All Projects → paraschopra → One Network Many Uses

paraschopra / One Network Many Uses

Four-in-one deep network: image search, image captioning, similar words and similar images using a single model

Projects that are alternatives of or similar to One Network Many Uses

Docs
MindSpore document
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Seq2seq tutorial
Code For Medium Article "How To Create Data Products That Are Magical Using Sequence-to-Sequence Models"
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Hep ml
Machine Learning for High Energy Physics.
Stars: ✭ 133 (+0%)
Mutual labels:  jupyter-notebook
Tensorflow realtime multi Person pose estimation
Multi-Person Pose Estimation project for Tensorflow 2.0 with a small and fast model based on MobilenetV3
Stars: ✭ 129 (-3.01%)
Mutual labels:  jupyter-notebook
Deep Reinforcement Learning In Trading
Stars: ✭ 129 (-3.01%)
Mutual labels:  jupyter-notebook
Deep Histopath
A deep learning approach to predicting breast tumor proliferation scores for the TUPAC16 challenge
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Attic Climate
Mirror of Apache Open Climate Workbench
Stars: ✭ 131 (-1.5%)
Mutual labels:  jupyter-notebook
Ghost Free Shadow Removal
[AAAI 2020] Towards Ghost-free Shadow Removal via Dual Hierarchical Aggregation Network and Shadow Matting GAN
Stars: ✭ 133 (+0%)
Mutual labels:  jupyter-notebook
Rwet
Notebooks and other materials for Reading and Writing Electronic Text
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Keras Mnist Tutorial
For a mini tutorial at U of T, a tutorial on MNIST classification in Keras.
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Huawei Digix Agegroup
2019 HUAWEI DIGIX Nurbs Solutions
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Security Api Solutions
Microsoft Graph Security API applications and services.
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Privateml
Various material around private machine learning, some associated with blog
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Google Colab Cloudtorrent
Colab Notebook Remote torrent client
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Algobook
A beginner-friendly project to help you in open-source contributions. Data Structures & Algorithms in various programming languages Please leave a star ⭐ to support this project! ✨
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Team Learning Nlp
主要存储Datawhale组队学习中“自然语言处理”方向的资料。
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook
Deeplearningbook Notes
Notes on the Deep Learning book from Ian Goodfellow, Yoshua Bengio and Aaron Courville (2016)
Stars: ✭ 1,672 (+1157.14%)
Mutual labels:  jupyter-notebook
Tensorflow In Practise Specialization
Four Courses Specialization Tensorflow in practise Specialization
Stars: ✭ 133 (+0%)
Mutual labels:  jupyter-notebook
Basketballvideoanalysis
Stars: ✭ 133 (+0%)
Mutual labels:  jupyter-notebook
Ml riskmanagement
Short Course - Applied Machine Learning for Risk Management
Stars: ✭ 132 (-0.75%)
Mutual labels:  jupyter-notebook

One neural network, many uses

Jupyter notebook corresponding to tutorial: One neural network network, many uses: image captioning, image search, similar images and similar words using one model

Requires following packages:

  • PyTorch
  • Numpy
  • Matplotlib

Dataset NOT included (please download separately)

To follow along, you'll need to download the Flickr8K dataset. Fill this form and you'll receive a download link on your email. (There's also a direct link to download the 1GB Flickr8K dataset, though not sure how long it'll stay like that). Extract the zip file in the 'Flicker8k_Dataset' folder in the same directory as your notebook. You'll also need to download captions from here. Extract captions in the folder 'caption_datasets'.

Example performance

Image caption example:

art 1

Similar words example

art 2

Similar images example

asrt 3

Image search example

search example

search example

Made by @paraschopra

MIT License.

Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].