@@ -42,16 +42,26 @@ Design and develop machine learning models that involve automatic speech recogni
Design and develop an appearance-based methodology with a multimodal convolutional neural network to evaluate the attention span with eye gaze estimation and head pose estimation
* Implement a real-time face identification method to recognize face and locate facial landmarks to categorize the facial features.
* Identify the eye gaze direction of the child when interacting with the assessment tasks by developing an eye gaze estimation model.
* Develop a head pose estimation model to identify the head pose of the child.
* Develop an algorithm to predict the attention scores base on the gaze estimation and head pose estimation models and derive average attention scores.
3. IT19170480 - Monarawila C.H.M.R.P.K
Design and develop a model which involves automatically drawn shapes and Sinhala digit recognition and analyzing with measuring completion time of each activity to analyze Sri Lankan aged 7-9 children's responses to scale their mental chronometry (processing speed) IQ capabilities.
* Classify Sinhala hand-written letters from children
* Classify hand-drawn shapes from children
* Classify shapes/symbols inside selected areas
4. IT19091044 - Gnanarathna E.D.K.V
Design and develop machine learning models that involve Automatic speech recognition with Pattern Matching of Transcripts and Error Analysis of verbal response, to assess and identify a detailed overview of Knowledge IQ of Sinhala speaking Sri Lankan Children between the ages 7-9.
* Develop a deep learning model for speech recognition of a Limited sinhala vocabulary in children.
* Develop a transcript matching algorithm to evaluate correct responses.
