14 papers
Show Me Examples: Inferring Visual Concepts from Image Sets
Nick Stracke, Kolja Bauer, Stefan Andreas Baumann +3
Vision-language models (VLMs) can follow complex textual instructions, yet they struggle to reason from purely visual context. In particular, current models fail to infer shared co…
RayDer: Scalable Self-Supervised Novel View Synthesis from Real-World Video
Ulrich Prestel, Stefan Andreas Baumann, Nick Stracke +1
Self-supervised novel view synthesis (NVS) remains challenging to scale, despite the abundance of video data, largely due to the brittleness of training on realistic videos and the…
Learning Long-term Motion Embeddings for Efficient Kinematics Generation
Nick Stracke, Kolja Bauer, Stefan Andreas Baumann +3
Understanding and predicting motion is a fundamental component of visual intelligence. Although modern video models exhibit strong comprehension of scene dynamics, exploring multip…
Envisioning the Future, One Step at a Time
Stefan Andreas Baumann, Jannik Wiese, Tommaso Martorella +2
Accurately anticipating how complex, diverse scenes will evolve requires models that represent uncertainty, simulate along extended interaction chains, and efficiently explore many…
Scalable High-Resolution Pixel-Space Image Synthesis with Hourglass Diffusion Transformers
Katherine Crowson, Stefan Andreas Baumann, Alex Birch +3
We present the Hourglass Diffusion Transformer (HDiT), an image generative model that exhibits linear scaling with pixel count, supporting training at high-resolution (e.g. $1024 \…
CleanDIFT: Diffusion Features without Noise
Nick Stracke, Stefan Andreas Baumann, Kolja Bauer +2
Internal features from large-scale pre-trained diffusion models have recently been established as powerful semantic descriptors for a wide range of downstream tasks. Works that use…