IITR Quantum Hackathon 2021
IITR Quantum Hackathon, organised by the Quantum Computing Group, IIT Roorkee and sponsored by IBM Quantum as a part of Qiskit Fall Fest, brought students with different backgrounds together from 16th-19th October 2021 to team up and build quantum computing projects using Qiskit. Brian Ingmanson from IBM Quantum conducted a workshop for the participants, giving an overview of Quantum Computing and IBM Quantum Labs with Qiskit. QCG IIT Roorkee covered some advanced concepts in Quantum Computing through interactive and hands-on workshops.
The virtual hackathon began with participants forming teams and pitching ideas on the Qiskit Fall Fest platform. All the teams were required to create GitHub repositories of their projects and present their work to the judges. In this blog, we will be sharing the results and experiences of the top three winning teams.
First Place - Quantum Maestros
Kamal Utla, Jezer Jojo, Sanskriti Sharan, Utkarsh Raj and Ankan Dey teamed up to develop a “Quantum Music Composer”. Their project can be viewed at https://quantummaestros.herokuapp.com.
We wanted to create a Quantum Music Composer that enabled the users to pick an emotion and get a composition tailored to that mood.
Our project mainly revolved around the quantum random number generators (QRNG). We started off by designing a circuit that was capable of picking preset chord progressions randomly for different parts of the composition. Then, we came up with a circuit to generate rhythms that our main melody followed and also created drum patterns for a separate category. So along with the different emotions that users could choose from, they now also had the option to pick “Quantum Jazz”, which uses random chord progressions from all the emotions and used an improvised drum pattern. For the rhythm generating circuit, we used 4 qubits with an Ry gate applied on each. What’s neat about using the Ry gates however, is that by using circuits for different drum elements with different gate parameters and different shot numbers, we can get drum beats that aren’t too far off from what one would hear in modern music, sounds exciting right? For melody, our first instinct was to treat it as a Markov process wherein each note depended on the note before it. Referring to a paper on Szegedy quantum walks on weighted graphs , we decided to consider the melody as a walk on a graph where each node is a note and each directed edge is assigned a weight - the probability of getting from one note to the other, and that’s our whole project.
Link to the project - https://github.com/jezerjojo14/QuantumMusicComposer
Second Place - Quanta Gambit
Kaustav Bhattacharjee, Jinal Shah, N. Arutkeerthi, K. Sudharshan and Bhavesh Gnnanapareddy teamed up to develop “Quantum Chess”.
This is a fun variant of chess where people can get a more intuitive sense of some aspects of quantum physics (like superposition and entanglement) and play without the aid of pre-existing chess theory. Pieces randomly change states like a quantum particle, making it a game of both wits and luck. Each piece is assigned two states it may possibly take. Perchance, if the two states are the same, the piece behaves like one in classical chess. After each move, the pieces have an equal probability of settling into either state. This collapse can be determined by a classical pseudorandom generator or by accessing the IBM quantum computer.
The randomize function helps us generate a random string of 32 bits, each bit corresponding to 32 squares. Depending on the value of 0 and 1 generated, the piece at that square will be decided. The random number generator basically applies Hadamard gates to every 32 qubits and measures each of these qubits with the quantum computer accessed by the IBM cloud. Another function, QRNG, is used to produce a 3-bit random number. This 3-bit QRNG is used to randomize the location of 5 pieces at the start of the game. Depending on the output decimal value of the 3-bit binary string, we associate the piece with being in superposition at the start of the game. Since we only need 5 numbers and a 3-bit decimal number can range from 0 to 7, we run a recursion over the loop whenever a decimal more significant than 4 is encountered in the generated string.
Link to the project - https://github.com/CtrlShiftSammy/Quantum-Chess
Third Place - QuaZers
Varun Srivatsav, Gargi Chandrakar, Masih Ahmed, Jatan Trivedi and Rohit Malawat teamed up to develop a model on “Quantum Image Processing using QPIE”.
Image processing is a rapidly growing, on-demand field which has its applications ranging from transmitting noiseless images and edge detection to medical imaging. Quantum Image Processing offers a great computational advantage over classical image processing techniques. In this project, we implemented, using Qiskit, the QPIE model of quantum image processing for processing quantum images. In this model, we encode the pixel values of an image into the probability amplitudes of a pure quantum state and their positions into computational basis states. This method has the proven advantage of using exponentially fewer qubits than its classical counterparts.
In this project, using QPIE modelling, we demonstrated the following: application of Fourier transform on quantum image, denoising and edge detection.
Link to the project - https://github.com/Quant02/Quatum-Image-Processing
Congratulations to all the winners and participants for producing some outstanding projects and strengthening the Qiskit community! Stay tuned for more such exciting events by QCG IIT Roorkee.
