NeuroPark - 3D Vehicle Detection
YOLOv8 + depth estimation + Open3D pipeline producing oriented 3D vehicle boxes from monocular input.
A computer vision project developed with Neurosoft that explores monocular perception for 3D understanding. The pipeline integrates object detection, depth estimation, point-cloud generation, and spatial alignment to produce usable vehicle scenes without relying on LiDAR as the primary sensor.
Validated oriented bounding boxes against LiDAR-based measurements
Reached roughly 5-6 seconds per high-resolution image for batch-style analysis
Presented the project at KPZ25 Engineering Conference after collaboration with Neurosoft
Impact
Validated monocular 2D-to-3D reconstruction for parking
Role
Team Lead / ML Engineer
Timeline
2025
Key tags
Problem
3D scene understanding is useful for autonomous parking, but richer sensors are expensive and not always available.
Solution
Built a pipeline that uses monocular RGB input, estimates depth, reconstructs point clouds, and localizes vehicles in aligned 3D world coordinates.
Architecture
YOLOv8 handles vehicle detection, depth estimation produces scene geometry, Open3D builds colored point clouds, and PCA/SVD-based geometry constructs oriented 3D boxes aligned with world coordinates.
Challenges
- Depth estimation uncertainty from single-camera input
- Spatial alignment and coordinate calibration across the full pipeline
- Balancing scene quality with processing speed in near-real-time scenarios
Technology stack
Continue exploring
More case studies
Lucid - AI Notes Platform
Helps students turn scattered lecture screenshots and notes into searchable study material.
Knowledge Graph RAG (ToPWR)
Helps students and staff query university information spread across pages, departments, and systems.
Medical Chatbot RAG
Provides educational health Q&A from retrieved source content instead of free-form LLM answers.