All Projects → basakbuluz → Visual Question Answering

basakbuluz / Visual Question Answering

Licence: mit
📷 ❓ Visual Question Answering Demo and Algorithmia API

Programming Languages

python
139335 projects - #7 most used programming language

Projects that are alternatives of or similar to Visual Question Answering

Vizwiz Vqa Pytorch
PyTorch VQA implementation that achieved top performances in the (ECCV18) VizWiz Grand Challenge: Answering Visual Questions from Blind People
Stars: ✭ 33 (+83.33%)
Mutual labels:  vqa, jupyter-notebook
Bottom Up Attention
Bottom-up attention model for image captioning and VQA, based on Faster R-CNN and Visual Genome
Stars: ✭ 989 (+5394.44%)
Mutual labels:  vqa, jupyter-notebook
Tbd Nets
PyTorch implementation of "Transparency by Design: Closing the Gap Between Performance and Interpretability in Visual Reasoning"
Stars: ✭ 345 (+1816.67%)
Mutual labels:  vqa, jupyter-notebook
Integer Programming With Python
Introduction to Linear and Integer Programming Book
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
Ilearndeeplearning.py
This repository contains small projects related to Neural Networks and Deep Learning in general. Subjects are closely linekd with articles I publish on Medium. I encourage you both to read as well as to check how the code works in the action.
Stars: ✭ 896 (+4877.78%)
Mutual labels:  jupyter-notebook
Ml unsam
Repositorio de la materia Aprendizaje Automático (2do cuatrimestre de 2020)
Stars: ✭ 18 (+0%)
Mutual labels:  jupyter-notebook
Tensorflow2 For Absolute Beginners
Stars: ✭ 18 (+0%)
Mutual labels:  jupyter-notebook
Simplest Tensorflow Tensorboard Mnist Embedding Visualisation
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
Jupyter Text2code
A proof-of-concept jupyter extension which converts english queries into relevant python code
Stars: ✭ 897 (+4883.33%)
Mutual labels:  jupyter-notebook
National generation capacity
Data package: national generation capacity
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
Algorithms
Repository for the Algorithms course
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
Computer Vision
Computer vision exercise with Python and OpenCV.
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
E Healthify
App developed for Hack-the-Crisis India.
Stars: ✭ 18 (+0%)
Mutual labels:  jupyter-notebook
Ansible Jupyter.dockerfile
Building the Docker image with Ansible and Jupyter.
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
Nlp Tutorial
Tutorial for Natural Language Processing
Stars: ✭ 18 (+0%)
Mutual labels:  jupyter-notebook
Uber profit model
A predictive model to help Uber drivers make more money
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
Symptoms Tracker
A scalable web app for tracking covid-19-related symptoms amongst confined populations
Stars: ✭ 18 (+0%)
Mutual labels:  jupyter-notebook
Dlsm
Deep Learning Study Material
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
Lab files
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook
Tensorflowmnist
Various neural networks on MNIST data using TensorFlow library
Stars: ✭ 17 (-5.56%)
Mutual labels:  jupyter-notebook

📷 ❓ Görsel Soru Cevaplama / Visual Question Answering

Görsel Soru Cevaplama sahip olduğumuz bir resim ile ilgili sorulan sorulara, resim içerisindeki bilgilerin analiz edilmesi ile cevaplar üretilmeye çalışılması problemi olarak tanımlanabilir.

Bu problemde metinler şeklinde ifade edilen soruların işlenmesi bir Doğal Dil İşleme problemi iken; resimler içerisinden cevapların üretiminde her bir soru ayrı bir Bilgisayarla Görü problemine işaret eder.

Visual Question Answering can be defined as the problem of trying to produce answers by analyzing the information in the picture.

In this problem, the questions expressed in the form of texts are a Natural Language Processing problem; each question in the production of answers within the pictures indicates a separate Computer Vision problem.

Genel olarak sisteme bakacak olursak:

If we look at the system in general:

alt text

Görsel soru cevaplama problemi için geliştirilen modellerin genel yaklaşımı :

The general approach of the models developed for the visual questioning problem is:

alt text


📌 Görsel Soru Cevaplama görevi için geliştirilen ve literatürdeki çalışmalarda sıklıkla kullanılan veri kümeleri

Data sets developed for the Visual Question Answering task and frequently used in studies in the literature


Blog 📝

Görsel soru cevaplama ile ilgili anlatım ve bu görev için sıklıkla kullanılan veri kümeleri hakkında detaylı bilgi edinmek için "Çok Gören Mi Bilir, Çok Soran Mı?" başlıklı blog yazıma göz atabilirsiniz.

You can browse my blog titled "Çok Gören Mi Bilir, Çok Soran Mı?" to get detailed information about the Visual Question Answering and the datasets frequently used for this task.


Uygulama (Implementation) 🔨

Gereksinimler (Requirements):

Demo 🖥️

Open In Colab

Bu jupyter notebook çalışma dosyası, verilen görüntü hakkında sorulan soruyu cevaplamak için önceden hazırlanmış modelleri kullanan basit bir Görsel Soru Cevaplama demosudur.

This jupyter notebook is a simple Visual Question answering demo that uses pretrained models to answer a given question about the given image.

API 💻

Geliştirdiğiniz ürünlere hızlı bir şekilde entegre edebileceğiniz API'yi kullanmak için buraya tıklayın.

Click here to use the API that you can integrate quickly into the products you have developed.

API Python Uygulama / API Python Implementation
Kurulum / Install

Algorithmia Python istemcisini pip ile yükleyin / Install the Algorithmia Python client with pip:

pip install algorithmia
Kullanım / Use
import Algorithmia

input = {
  "image": "data://yavuzkomecoglu/DL_VQA/test.jpg",
  "question": "What vehicle is in the picture?"
}
client = Algorithmia.client('YOUR_API_KEY')
algo = client.algo('yavuzkomecoglu/VQA/0.1.1')
print(algo.pipe(input).result)

Örnek Tahminler / Sample predictions

VQA modeli tarafından tahmin edilen bazı cevaplar.

Some answers predicted by the VQA model.

Q: How is the weather? A: Sunny! (%97.23)

Q: How many girls are in the picture? A: 2! (%61.98)

Q: What is done in the picture? A: Surfing! (%99.43)

Q: What does the sign say? A: Stop! (%28.61)


Referanslar / References

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].