About Me
Hi! I am Abhishek Reddy Malreddy, a graduate student pursuing my M.S. in Artificial Intelligence Engineering at Carnegie Mellon University. My expertise lies in leveraging AI and ML to solve complex engineering challenges.
Projects
Segmentation Road Scene Dataset in Adverse Weather Conditions
Developed the IDD-AW dataset, the most relevant dataset for semantic scene understanding in Indian driving scenarios under adverse weather conditions. Benchmarked state-of-the-art semantic segmentation models and introduced the "Safe mean Intersection over Union (Safe mIoU)" metric to penalize dangerous mispredictions.
The image illustrates predictions made on the IDD-AW test set using pretrained models from ACDC, IDD, and IDDAW, along with ground truth and severity maps generated by the IDD-AW pretrained model’s predictions. In the severity maps, colors signify various danger levels, where yellow indicates misclassification at Level 3 (the lowest level of the tree), orange represents misclassification at Level 2, and red corresponds to Level 1, collectively indicating the overall danger level of the driving scene.
Optimizing Vehicle Traversal Time Through RL-Based Lane Selection Strategy
Improved vehicle traversal efficiency by optimizing lane-level dynamics and utilizing vehicle-to-vehicle communication. Conducted experiments in simulated congested traffic scenarios using SUMO software.
Reinforcement Learning based Lane Search (RLLS) Algorithm
Other Projects
- Data Analysis and Machine Learning Modeling on FIFA Dataset using PySpark and SQL on GCP.
- Developed deep learning-based steganography techniques to securely embed one image within another.
- Utilized Denoising Diffusion Probabilistic Models to generate new image samples for the Amsterdam Library of Textures Dataset and the MNIST Dataset.
Research & Work Experience
Machine Learning Lab, IIIT Hyderabad & DRDO
Research Assistant
May 2023 – July 2024
Collaborated on research focused on safety and robustness in autonomous navigation. Developed Indian Driving Scene Dataset for unstructured traffic and adverse weather scenarios. Worked on image enhancement with diffusion models and GANs, NIR+RGB fusion, and segmentation.
Genpact India
Management Trainee
August 2022 – April 2023
Managed the incident ticket database and ensured client SLA compliance.
Pebble DLT
Developer
May 2020 – October 2020
Built a decentralized Pebble Network for transaction management, involving timestamp propagation, node categorization, and queue management using unique public key identifiers.
Education
Carnegie Mellon University
M.S. in Artificial Intelligence Engineering - Materials Science and Engineering (Aug 2024 - Dec 2025 expected)
GPA: 4.0/4.0
NIT Calicut
Bachelor of Technology in Mechanical Engineering (Jun 2018 - May 2022)
GPA: 3.45/4.0
Coursework
-
Systems & Tool Chains for AI Engineers
-
Machine Learning & Artificial Intelligence for Engineers
-
Computer Vision
-
Learning for 3D Vision
-
Integrated Intelligence in Robotics: Vision Language Planning
-
Methods of Computational Materials Science
-
Structure and Characterization of Materials
Publications
F. A. Shaik, A. Reddy, N. R. Billa, K. Chaudhary, S. Manchanda, and G. Varma, "IDD-AW: A Benchmark for Safe and Robust Segmentation of Drive Scenes in Unstructured Traffic and Adverse Weather," in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Jan. 2024, pp. 4614–4623.
Skills
- Programming Skills: Python, C, C++, SQL, Git
- Machine Learning Tools: PyTorch, TensorFlow, Keras, Scikit-learn, OpenCV, Reinforcement Learning
- Technologies: GCP, Apache Spark, Docker, Data Modeling, Data Pipelines, MLOps, Eclipse SUMO, LaTeX
- Soft Skills: Leadership, Team Collaboration, Communication, Research
- Extracurriculars: Cricket (Represented NIT Calicut Team), Farming (Working on technology adoption in Agriculture)
Certifications
Accomplishments
-
Design and Analysis of Air Conditioning System: Conducted research on enhancing air efficiency with increased oxygen and purity levels in critical care hospitals, NIT Calicut.
Supervisor: Dr. Biju T Kuzhiveli (2021-22) -
Summer Research Internship: Participated in SRISHTI, IIIT Hyderabad, focusing on technological innovations. (2022)
-
Caterpillar IDP Case Challenge: Ranked among the top 5 finalists out of 724 teams across India at Shaastra 2021, IIT Madras.
Addressed the problem of reducing frictional and parasitic losses in heavy-duty diesel engines. (2021)