Master's student. Deep learning, machine learning, and computer vision.
Explore the application of probabilistic machine learning in predicting house prices, highlighting how accounting for uncertainty leads to models that better reflect real-world variability compared to deterministic approaches.
Multi-View Stereo (MVS) has become a key player in fields like 3D mapping, video games, and autonomous driving, allowing us to create detailed 3D reconstructions from multiple images. Recent advancements, especially those leveraging deep learning, have made MVS techniques more efficient and accurate. In this blog post, we'll dive into the latest methods shaping the future of MVS, with a particular focus on MVSNet and its successors. I'll break down how MVSNet works, covering essential steps like feature extraction, cost volume regularization, and reconstruction. At the end, there is a comparison how various models perform, using popular benchmarks like the DTU and Tanks and Temples datasets.