Ivo Danihelka
Papers
1
Total Citations
22
H-Index
1
About
Ivo Danihelka is a research scientist whose work lies at the intersection of deep learning, probabilistic modeling, and video generation. He is best known for his pioneering contributions to autoregressive pixel-level video models, most notably through the introduction of the Video Pixel Network (VPN). In this highly cited 2016 paper, Danihelka and his co-authors proposed a novel probabilistic framework that estimates the joint distribution of raw pixel values across time, space, and color channels, structuring the neural architecture to reflect the four-dimensional dependencies inherent in video tensors. This work, which has accumulated over 22 citations, laid foundational groundwork for subsequent advances in video generation and predictive modeling. Danihelka’s research exemplifies a rigorous approach to capturing complex temporal and spatial correlations, pushing the boundaries of what generative models can achieve with raw visual data. His contributions continue to influence researchers working on video synthesis, probabilistic deep learning, and structured autoregressive models.
Research Focus
Key Achievements
Top Papers
- 1Video Pixel Networks22 citations · 2016