Michael Oechsle
YOU?
Author Swipe
View article: AnyUp: Universal Feature Upsampling
AnyUp: Universal Feature Upsampling Open
We introduce AnyUp, a method for feature upsampling that can be applied to any vision feature at any resolution, without encoder-specific training. Existing learning-based upsamplers for features like DINO or CLIP need to be re-trained for…
View article: Learning Neural Exposure Fields for View Synthesis
Learning Neural Exposure Fields for View Synthesis Open
Recent advances in neural scene representations have led to unprecedented quality in 3D reconstruction and view synthesis. Despite achieving high-quality results for common benchmarks with curated data, outputs often degrade for data that …
View article: CubeDiff: Repurposing Diffusion-Based Image Models for Panorama Generation
CubeDiff: Repurposing Diffusion-Based Image Models for Panorama Generation Open
We introduce a novel method for generating 360° panoramas from text prompts or images. Our approach leverages recent advances in 3D generation by employing multi-view diffusion models to jointly synthesize the six faces of a cubemap. Unlik…
View article: Gaussians-to-Life: Text-Driven Animation of 3D Gaussian Splatting Scenes
Gaussians-to-Life: Text-Driven Animation of 3D Gaussian Splatting Scenes Open
State-of-the-art novel view synthesis methods achieve impressive results for multi-view captures of static 3D scenes. However, the reconstructed scenes still lack "liveliness," a key component for creating engaging 3D experiences. Recently…
View article: Evolutive Rendering Models
Evolutive Rendering Models Open
The landscape of computer graphics has undergone significant transformations with the recent advances of differentiable rendering models. These rendering models often rely on heuristic designs that may not fully align with the final render…
View article: Splat-SLAM: Globally Optimized RGB-only SLAM with 3D Gaussians
Splat-SLAM: Globally Optimized RGB-only SLAM with 3D Gaussians Open
3D Gaussian Splatting has emerged as a powerful representation of geometry and appearance for RGB-only dense Simultaneous Localization and Mapping (SLAM), as it provides a compact dense map representation while enabling efficient and high-…
View article: RadSplat: Radiance Field-Informed Gaussian Splatting for Robust Real-Time Rendering with 900+ FPS
RadSplat: Radiance Field-Informed Gaussian Splatting for Robust Real-Time Rendering with 900+ FPS Open
Recent advances in view synthesis and real-time rendering have achieved photorealistic quality at impressive rendering speeds. While Radiance Field-based methods achieve state-of-the-art quality in challenging scenarios such as in-the-wild…
View article: Differentiable Optimization for Orchestration: Resource Offloading for Vehicles in Smart Cities
Differentiable Optimization for Orchestration: Resource Offloading for Vehicles in Smart Cities Open
Connected and Autonomous Vehicles (CAV) which interact with Roadside Units (RSU) as part of a smart city infrastructure are currently seeing first real-world deployments. Not only can CAVs benefit from access to a cities’ infrastructure by…
View article: UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction
UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction Open
Neural implicit 3D representations have emerged as a powerful paradigm for reconstructing surfaces from multi-view images and synthesizing novel views. Unfortunately, existing methods such as DVR or IDR require accurate per-pixel object ma…
View article: UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction
UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for\n Multi-View Reconstruction Open
Neural implicit 3D representations have emerged as a powerful paradigm for\nreconstructing surfaces from multi-view images and synthesizing novel views.\nUnfortunately, existing methods such as DVR or IDR require accurate per-pixel\nobject…
View article: Learning Implicit Surface Light Fields
Learning Implicit Surface Light Fields Open
Implicit representations of 3D objects have recently achieved impressive results on learning-based 3D reconstruction tasks. While existing works use simple texture models to represent object appearance, photo-realistic image synthesis requ…
View article: Differentiable Volumetric Rendering: Learning Implicit 3D Representations Without 3D Supervision
Differentiable Volumetric Rendering: Learning Implicit 3D Representations Without 3D Supervision Open
Learning-based 3D reconstruction methods have shown impressive results. However, most methods require 3D supervision which is often hard to obtain for real-world datasets. Recently, several works have proposed differentiable rendering tech…
View article: Texture Fields: Learning Texture Representations in Function Space
Texture Fields: Learning Texture Representations in Function Space Open
In recent years, substantial progress has been achieved in learning-based reconstruction of 3D objects. At the same time, generative models were proposed that can generate highly realistic images. However, despite this success in these clo…
View article: Occupancy Networks: Learning 3D Reconstruction in Function Space
Occupancy Networks: Learning 3D Reconstruction in Function Space Open
With the advent of deep neural networks, learning-based approaches for 3D reconstruction have gained popularity. However, unlike for images, in 3D there is no canonical representation which is both computationally and memory efficient yet …