Publications by Sabine Süsstrunk

FDS: Frequency-Aware Denoising Score for Text-Guided Latent Diffusion Image Editing

Text-guided image editing using Text-to-Image (T2I) models often fails to yield satisfactory results, frequently introducing unintended modifications, such as the loss of local detail and color changes. In this paper, we analyze these failure cases and att ...

IEEE2025

Dark Noise Diffusion: Noise Synthesis for Low-Light Image Denoising

Sabine Süsstrunk, Liying Lu

Low-light photography produces images with low signal-to-noise ratios due to limited photons. In such conditions, common approximations like the Gaussian noise model fall short, and many denoising techniques fail to remove noise effectively. Although deep- ...

2025

VolRecon: Volume Rendering of Signed Ray Distance Functions for Generalizable Multi-View Reconstruction

Sabine Süsstrunk, Yufan Ren, Tong Zhang

The success of the Neural Radiance Fields (NeRF) in novel view synthesis has inspired researchers to propose neural implicit scene reconstruction. However, most existing neural implicit reconstruction methods optimize perscene parameters and therefore lack ...

2024

Efficient Temporally-Aware DeepFake Detection using H.264 Motion Vectors

Sabine Süsstrunk, Alessio Verardo, Yufan Ren, Peter Arpad Grönquist, Qingyi He

Video DeepFakes are fake media created with Deep Learning (DL) that manipulate a person’s expression or identity. Most current DeepFake detection methods analyze each frame independently, ignoring inconsistencies and unnatural movements between frames. Som ...

2024

Mitigating Object Dependencies: Improving Point Cloud Self-Supervised Learning through Object Exchange

Mathieu Salzmann, Sabine Süsstrunk, Tong Zhang, Yi Wu

In the realm of point cloud scene understanding, particularly in indoor scenes, objects are arranged following human habits, resulting in objects of certain semantics being closely positioned and displaying notable inter-object correlations. This can creat ...

2024

Exploiting the Signal-Leak Bias in Diffusion Models

Sabine Süsstrunk, Radhakrishna Achanta, Mahmut Sami Arpa, Athanasios Fitsios, Martin Nicolas Everaert

There is a bias in the inference pipeline of most diffusion models. This bias arises from a signal leak whose distribution deviates from the noise distribution, creating a discrepancy between training and inference processes. We demonstrate that this signa ...

2024

InNeRF360: Text-Guided 3D-Consistent Object Inpainting on 360° Neural Radiance Fields

Sabine Süsstrunk, Tong Zhang, Dongqing Wang

We propose InNeRF360, an automatic system that accurately removes text-specified objects from 360 • Neural Radiance Fields (NeRF). The challenge is to effectively remove objects while inpainting perceptually consistent content for the missing regions, whic ...

IEEE2024

AdaNCA: Neural Cellular Automata as Adaptors for More Robust Vision Transformer

Sabine Süsstrunk, Yitao Xu, Tong Zhang

Vision Transformers (ViTs) demonstrate remarkable performance in image classification through visual-token interaction learning, particularly when equipped with local information via region attention or convolutions. Although such architectures improve the ...

2024

Mesh Neural Cellular Automata

Sabine Süsstrunk, Ehsan Pajouheshgar, Yitao Xu, Tong Zhang

Texture modeling and synthesis are essential for enhancing the realism of virtual environments. Methods that directly synthesize textures in 3D offer distinct advantages to the UV-mapping-based methods as they can create seamless textures and align more cl ...

Association for Computing Machinery (ACM)2024

NoiseNCA: Noisy Seed Improves Spatio-Temporal Continuity of Neural Cellular Automata

Sabine Süsstrunk, Ehsan Pajouheshgar, Yitao Xu

Neural Cellular Automata (NCA) is a class of Cellular Automata where the update rule is parameterized by a neural network that can be trained using gradient descent. In this paper, we focus on NCA models used for texture synthesis, where the update rule is ...

MIT Press2024

Fast Adversarial Training With Adaptive Step Size

Mathieu Salzmann, Sabine Süsstrunk, Jue Wang, Chen Liu, Yong Zhang, Zhuoyi Huang

While adversarial training and its variants have shown to be the most effective algorithms to defend against adversarial attacks, their extremely slow training process makes it hard to scale to large datasets like ImageNet. The key idea of recent works to ...

Ieee-Inst Electrical Electronics Engineers Inc2023

Diffusion in Style

Sabine Süsstrunk, Radhakrishna Achanta, Mahmut Sami Arpa, Martin Nicolas Everaert

We present Diffusion in Style, a simple method to adapt Stable Diffusion to any desired style, using only a small set of target images. It is based on the key observation that the style of the images generated by Stable Diffusion is tied to the initial lat ...

2023

Vision Transformer Adapters for Generalizable Multitask Learning

Mathieu Salzmann, Sabine Süsstrunk, Deblina Bhattacharjee

We introduce the first multitasking vision transformer adapters that learn generalizable task affinities which can be applied to novel tasks and domains. Integrated into an off-the-shelf vision transformer backbone, our adapters can simultaneously solve mu ...

Ieee Computer Soc2023

TempSAL - Uncovering Temporal Information for Deep Saliency Prediction

Mathieu Salzmann, Sabine Süsstrunk, Bahar Aydemir, Tong Zhang

Deep saliency prediction algorithms complement the object recognition features, they typically rely on additional information such as scene context, semantic relationships, gaze direction, and object dissimilarity. However, none of these models consider th ...

Ieee Computer Soc2023

Saliency prediction in 360° architectural scenes: Performance and impact of daylight variations

Marilyne Andersen, Sabine Süsstrunk, Kynthia Chamilothori, Bahar Aydemir, Caroline Karmann, Seungryong Kim

Saliency models are image-based prediction models that estimate human visual attention. Such models, when applied to architectural spaces, could pave the way for design decisions where visual attention is taken into account. In this study, we tested the pe ...

2023

Dense Multitask Learning to Reconfigure Comics

Mathieu Salzmann, Sabine Süsstrunk, Deblina Bhattacharjee

In this paper, we develop a MultiTask Learning (MTL) model to achieve dense predictions for comic panels to, in turn, facilitate the transfer of comics from one publication channel to another by assisting authors in the task of reconfiguring their narrativ ...

IEEE2023

Spatiotemporal Self-supervised Learning for Point Clouds in the Wild

Mathieu Salzmann, Sabine Süsstrunk, Tong Zhang

Self-supervised learning (SSL) has the potential to ben- efit many applications, particularly those where manually annotating data is cumbersome. One such situation is the semantic segmentation of point clouds. In this context, ex- isting methods employ co ...

IEEE2023

DyNCA: Real-time Dynamic Texture Synthesis Using Neural Cellular Automata

Sabine Süsstrunk, Ehsan Pajouheshgar, Yitao Xu, Tong Zhang

Current Dynamic Texture Synthesis (DyTS) models can synthesize realistic videos. However, they require a slow iterative optimization process to synthesize a single fixed-size short video, and they do not offer any post-training control over the synthesis p ...

Ieee Computer Soc2023

VETIM: Expanding the Vocabulary of Text-to-Image Models only with Text

Sabine Süsstrunk, Radhakrishna Achanta, Mahmut Sami Arpa, Martin Nicolas Everaert

Text-to-image models, such as Stable Diffusion, can generate high-quality images from simple textual prompts. With methods such as Textual Inversion, it is possible to expand the vocabulary of these models with additional concepts, by learning the vocabula ...

BMVA2023

Towards Stable and Efficient Adversarial Training against $l_1$ Bounded Adversarial Attacks

Mathieu Salzmann, Sabine Süsstrunk, Chen Liu, Zhuoyi Huang, Yulun Jiang

We address the problem of stably and efficiently training a deep neural network robust to adversarial perturbations bounded by an

l_1

norm. We demonstrate that achieving robustness against

l_1

-bounded perturbations is more challenging than in the

l_2

...

2023