Digitizing the Documents and Books for Accessibility With Machine Translation and NLP

Digitizing the Documents and Books for Accessibility With Machine Translation and NLP

Develop a machine translation model that automatically translates the given documents written in Bhutan’s local language to English and creates a digital copy by storing it in the cloud. In this 8-week challenge, you will join a collaborative team of 50 AI engineers from all around the world.

 

The idea

In this first-of-its-kind virtual ideathon in collaboration with Druk Holding & Investments (the commercial arm of the Royal Government of Bhutan), Machine Learning and Data Science enthusiasts had the opportunity to submit their ideas to use technology to help tackle climate change, healthcare, energy, and other social good challenges in Bhutan.

The aim of the ideathon was to work on the most impactful idea through the Omdena 8-week challenge to make a change in Bhutan. The author of the winning solution is Akash Phaniteja Nellutla.

 

The problem

In Bhutan, eighteen different languages are spoken and of those only Dzongkha has a native literary tradition. Digitization and translation of Bhutanese literature to English will help in the globalization and preservation of the literature with the flexibility, versatility, and richness of English.

Bhutan’s development has been guided by its philosophy of gross national Happiness and focuses heavily on its economic growth and development. With a program to digitize texts and documents related to cultural heritage and literature, it has the capacity to encourage the development of its digital cultural heritage, creative industries, and IT. 

More than two thousand temples, and monasteries (government and private) – Expected to hold important text, documents, and manuscripts. With a program to digitize texts and documents related to cultural heritage and literature, it has the capacity to encourage the development of its digital cultural heritage, creative industries, and IT.

 

The project goals

“Conserving the past for the future”.

The application of NLP such as machine translation would benefit the tourists to interact with the local people of Bhutan to get more insights into the culture and tradition.

The model can be also expanded to other languages if required.

 

The data

Data provided by the partner will include a list of official documents and old literature.

This challenge will also include data collection.

 

Why join? The uniqueness of Omdena AI Challenges

A collaborative experience you never had in your working life! For the next eight weeks, you will not only build AI solutions to make a real-world impact but also go through an entire data science project lifecycle. This covers problem scoping, data collection, and preparation, as well as modeling for deployment.

And the best part is that you will join a global and collaborative team of changemakers. Omdena AI Challenges are not a competition or hackathon but a real-world project that will take your experience of what is possible through collaboration to a new level.

 

Find more information on how an Omdena project works

 

Using Image Analysis to Estimate the Density of Blood Cells

Using Image Analysis to Estimate the Density of Blood Cells

Join a global team of 50 AI changemakers in this high-impact 2-month challenge to estimate the density of blood cells

This challenge requires experience in Image Analysis, Data Analysis, and Machine Learning.

You will join a team of 50 technology changemakers from different experience levels and backgrounds.

 

The Context

LiteBC is developing the first non-invasive blood count analyzer, the HemoScope, that will enable obtaining blood count results within 30 seconds, in a non-invasive and affordable device. The HemoScope will impact the health of society, enabling remote healthcare and reducing financial barriers for the lower social-economic populations. LiteBC will acquire huge amounts of data from the HemoScope, which will be in video format, consisting of blood vessel images from two distinct microscopes. Each subject’s data will be accompanied with ground truth data – taken using the traditional blood count analyzers in the hospital.

During the challenge the videos/images will be analyzed and transformed into single numbers representing the quantities of various parameters: white blood cells count, red blood cells count, hemoglobin levels, etc.

 

The Project Goals

The AI solution should be able to estimate the density of various blood cell types based on visual data acquired by our device. The estimation can, for example, be constructed as a segmentation problem followed by a counting process of single blood cells. The training and testing processes will be based on both annotated data and results of traditional blood tests.

The deliverables should be a procedure to estimate different parameters of the blood count, based on our visual data.

 

Why join? The uniqueness of Omdena AI Challenges

A collaborative experience you never had in your working life! For the next eight weeks, you will not only build AI solutions to make a real-world impact but also go through an entire data science project lifecycle. This covers problem scoping, data collection, and preparation, as well as modeling for deployment.

And the best part is that you will join a global and collaborative team of changemakers. Omdena AI Challenges are not a competition or hackathon but a real-world project that will take your experience of what is possible through collaboration to a new level.

 

Find more information on how an Omdena project works

 

NLP and Conversational AI for Compassionate AI Psychologist with Human-like Memory

NLP and Conversational AI for Compassionate AI Psychologist with Human-like Memory

Build an AI psychologist which will be able to conduct compassionate conversations by using a voice-based conversational AI approach, with the ability to identify the emotions of the users and their underlying mental health state. In this 8-week challenge, you will join a collaborative team of 50 AI engineers.

 

The problem

During the pandemic, we saw a mental health ‘surge’ with a sharp increase in people experiencing anxiety (health anxiety, loneliness, etc.) and stress (including Post Traumatic Stress).

Surveys by such organizations as YouGov and Mental Health America show that, during the pandemic, workers were more stressed than ever. This included anxiety regarding health, finance, and employment, as well as broader issues associated with PTSD, burnout, and moderate depression.

Nearly two-thirds of the UK population (63%) felt anxious at least several times a month during the Spring of 2020. One in five (20%) reported feeling anxious on most days of the week or even more frequently.

Mental Health America’s 2021 study shows the strain on employees, with burnout prevalent and more than half of the respondents actively looking for another job.

There is an increased strain on an already-overstretched mental health service that, in some part, has been accommodated by a growing number of apps as well as telehealth calls (Skype / Zoom calls to human therapists).

 

The project goals

The project goal is to build an AI psychologist which will be able to conduct compassionate conversations by using a voice-based conversational AI approach, with the ability to identify the emotions of the user and their underlying mental health state and the capacity to remember these users’ states by possessing a human-like memory.

 

The data

This project will require the manual collection of data.

 

Why join? The uniqueness of Omdena AI Challenges

A collaborative experience you never had in your working life! For the next eight weeks, you will not only build AI solutions to make a real-world impact but also go through an entire data science project lifecycle. This covers problem scoping, data collection, and preparation, as well as modeling for deployment.

And the best part is that you will join a global and collaborative team of changemakers. Omdena AI Challenges are not a competition or hackathon but a real-world project that will take your experience of what is possible through collaboration to a new level.

 

Find more information on how an Omdena project works

 

Violence Detection Between Children and Caregivers Using Computer Vision

Violence Detection Between Children and Caregivers Using Computer Vision

A team of 50 Omdena AI changemakers collaborated with Israel-based startup EyeKnow AI to apply deep learning to build a computer vision model for violence detection. The model can help not only detect but in the future also prevent violent behaviour applied to children by caregivers.

 

The problem

Child maltreatment presents a substantial public health concern. Estimates using Child Protective Service (CPS) reports from the National Child Abuse and Neglect Data System (NCANDS) suggest that 678,810 youth were subjected to maltreatment in 2012, with 18% of these experiencing physical abuse (). Additionally, a large proportion of cases are undetected by CPS, suggesting that more youth are likely subjected to abusive or neglectful behavior (). Most seriously, maltreatment was responsible for an estimated 1,640 youth fatalities in 2012 ().

 

The project outcomes 

The data

Two datasets, one is a caregiver-to-senior violence dataset, made out of 500 clips sourced entirely from YouTube. The 2nd dataset comprises 500 clips of caregiver-to-child aggression/violence, driven by YouTube clips and unique data obtained through partnerships with EyeKnow’s partners. 

 

The machine learning models

The contributors of the challenge defined several approaches to build a model to detect violent interaction or any relevant interaction between the entities (caregivers, elderly, children). The first step of this approach was to see the entities, which the team did by utilizing object detection.

The team applied frame-level entity annotation to label the caregivers, children, and elderly. After this step, the collaborators trained an object detection model and implemented an ML pipeline. This pipeline ingests video recordings from CCTV or other sources and outputs frame-level information about the number and type of entities on the frame level. In addition, bounding box-based overlap analysis was included in the pipeline, which flags frames that potentially contain interaction of high intensity (potentially violent). 

Next to this pipeline, the team applied video classification modeling utilizing deep neural networks. This approach combined pre-trained models for feature extraction with sequence modeling to capture temporal relationships. 

All the developed models and approaches run in a Python application. The application is highly modular and serves multiple purposes. By modifying a configuration file (parameters JSON file), the user can execute training of component models or manage inference and process video files.

 

Digitizing Floor Plan Layouts using AI

Digitizing Floor Plan Layouts using AI

50 AI engineers collaborated in this high-impact 2-months innovation project to identify and construct digital objects from floor plans using computer vision.

 

The problem

To recognize floor plan elements in a layout requires manual labor to draw the different elements over the image. The goal of this project has been to improve the efficiency of this manual effort by automatically identifying the relevant types of objects present using state-of-the-art deep learning and computer vision approaches.

 

The project outcomes

The team built computer vision models to digitize the floor plan from architectural blueprints. The team successfully applied the following methods in achieving the tasks:

  • Object detection,
  • Image segmentation using Mask RCNN
  • Improved Optical character recognition (OCR) using the provided datasets,
  • Identifying languages other than English on floor plans

 

Data

Archilyse provided a large set of bitmap images of different sizes and dimensions, along with the bounding boxes of the relevant type of elements manually drawn. Examples of those are walls, columns, railings, kitchen furniture, shower, windows, doors, bathtub, bedroom area, kitchen area, etc.