Dennis Slobodzian

Data Scientist • Computer Vision / Machine Learning

• Develop and validate production-oriented object-detection and segmentation models through controlled experiments, debugging, data-quality analysis, and deployment-focused evaluation.
• Customized a YOLOX architecture for proprietary imagery, delivering approximately 2× faster inference; built automated mask-annotation tooling for large cell images.

Computer Vision / Machine Learning Intern

• Optimized an annotation and model-development data-aggregation workflow from more than 24 hours to under 2 hours; developed annotation and mask-generation automation for high-resolution imagery.
• Developed artificial annotations with GANs and VAEs to improve data coverage in limited datasets.
• Created synthetic images emplacing objects from positive signal data to negative signal data with Poisson blending to improve detector performance.

Robotics Software & Controls Mentor

• Develop and mentor C++ control software for mobile competition robots, integrating drivetrain and mechanism control, motor controllers, sensors, autonomous behaviors, telemetry, and physical-hardware validation.
• Use AprilTag measurements and Kalman-filter-based pose estimation for localization and autonomous operation; implement and teach LQR/state-space feedback for differential-drive control.
• Supported migration of robot-control code from Java to C++ on constrained Xilinx hardware; profiled and debugged control-loop performance, with measured runtime improving from approximately 20 ms to 1–2 ms.

• Built a Drake simulation experiment using a KUKA iiwa arm and V-JEPA2 embeddings to test whether pretrained video representations capture robot-motion structure. Evaluated motion-class separation with nearest-neighbor accuracy, clustering metrics, and cosine similarity, then used latent similarity to derive a dense reach-task progress signal.

• Built a modular YOLO-based 3D vision pipeline for NVIDIA Jetson; implemented CUDA preprocessing and TensorRT-optimized inference for real-time edge deployment. Added tests, CMake builds, Docker assets, and GitHub Actions.

• Designed a motor-controller PCB in Altium and a wireless 8 kHz mouse/dongle firmware project, applying schematic/PCB design, embedded timing, hardware interfaces, debugging, and hardware/software integration.

B.S., Electrical & Computer Engineering

Honors: Academic Excellence in Computer Science; Robotics Award

Slobodzian, D., Kordijazi, A. “Deep Learning Framework for Early Detection of Pancreatic Cancer Using Multi-modal Medical Imaging Analysis.” Journal of Imaging Informatics in Medicine, 2026. doi:10.1007/s10278-026-01998-w