Page 1

Image annotation and classification using deep learning models



Classifying and annotating images is an important task in machine learning. Many algorithms have been proposed for these tasks, based on features such as color, texture, and shape. The success of these algorithms is dependent on the selection of features. Deep learning models are widely used to learn abstract, high-level representations from raw data. Energy-based models are the most commonly used deep learning models formed by pre-training the individual restricted Boltzmann machines in a layerwise fashion and then stacking together and training them using error backpropagation. In the deep convolutional neural networks, the convolution operation is used to extract features from different sub-regions of the images to learn better representations. To reduce the time taken for training, models that use convex optimization and kernel trick have been proposed. Continue Reading →


Affective Mario

Emotion is defined as a natural instinctive state of mind deriving from one’s circumstances, mood, or relationships with others. Emotions are believed to be species specific rather than culture-specific. In the case of humans, emotions are expressed through beha8567090003_100e228021_ovior, actions, thoughts and feelings. Among these expressions facial expressions is one of the most natural forms of display of human emotions. Facial expressions in humans are controlled by the action of more than 40 muscles. A motion detector, such as Kinect, used for gaming can track the movements of these muscles. Using a machine learning technique, these movements can be classified as different emotions.

Continue Reading →


Community Detection in Large Social Networks Using PCCA+


Communities in the American Football Club dataset

Community detection can reflect overall structure of a network and thus it can help for many real case scenarios like product marketing. To extract such communities from network one uses the objective function which captures the intuition of communities with high intra-community edges and fewer inter-community edges. Since optimizing such objective function is NP-hard, many researchers have tried many heuristics to find approximate communities. Most of the community detection algorithms based on greedy algorithms perform poorly on large complex networks. Moreover, many algorithms for community detection also require some prior knowledge of the community structure, e.g., the number of the communities, which is very difficult to be obtained in real-world networks. Algorithms are evaluated based on score called modularity. The algorithm which maximizes such modularity score is considered to be the best algorithm. In this project, we propose a new community detection algorithm which is based on Perron Cluster Analysis. Continue Reading →


Apocalypse: Evasion of the Wipeout




I am excited to tell you that we have come up with our new game.

We used Blender for all 3D modelling for Apocalypse. For our game engine, we chose to work with BGE and added our own scripts in Python to extend it when required.

The game takes you to a post-apocalyptic world where every race starts to doubt and hate the other. The protagonist, in his thrilling journey, comes across several clues, one leading to the other and tries to uncover the truth behind the destruction. Continue Reading →


Demo of Official website VSSUT

After spending close to 10 days redesigning the look and feel and functionality of the proposed official website of VSSUT, finally we came up with something entirely new. This demonstration shows a screen-cast of the presentation I gave in the AVC to the Vice Chancellor and the Professors.


  • Niharjyoti Sarangi
  • Himanshu Patel
  • Piyush Mishra
  • Soumyaranjan Mohanty


Thank you all for the effort and dedication.