4 papers
Pixel-to-4D: Camera-Controlled Image-to-Video Generation with Dynamic 3D Gaussians
Melonie de Almeida, Daniela Ivanova, Tong Shi +2
Humans excel at forecasting the future dynamics of a scene given just a single image. Video generation models that can mimic this ability are an essential component for intelligent…
Neural 3D Reconstruction of Planetary Surfaces from Descent-Phase Wide-Angle Imagery
Melonie de Almeida, George Brydon, Divya M. Persaud +2
Digital elevation modeling of planetary surfaces is essential for studying past and ongoing geological processes. Wide-angle imagery acquired during spacecraft descent promises to…
Unsupervised Segmentation by Diffusing, Walking and Cutting
Daniela Ivanova, Marco Aversa, Paul Henderson +1
We propose an unsupervised image segmentation method using features from pre-trained text-to-image diffusion models. Inspired by classic spectral clustering approaches, we construc…
ARTeFACT: Benchmarking Segmentation Models on Diverse Analogue Media Damage
Daniela Ivanova, Marco Aversa, Paul Henderson +1
Accurately detecting and classifying damage in analogue media such as paintings, photographs, textiles, mosaics, and frescoes is essential for cultural heritage preservation. While…