-
University of Canterbury
- Christchurch, New Zealand
- https://ucvision.org.nz/
Stars
Official Implementation of Upsample Anything: A Simple and Hard to Beat Baseline for Feature Upsampling
Efficient Universal Perception Encoder: a single on-device vision encoder with versatile representations that match or exceed specialized experts across multiple task domains.
[ICLR '26 Oral] Official repository of the paper "AnyUp: Universal Feature Upsampling".
【RA-L 2026】 A Robust and Efficient LiDAR-Inertial Odometry System with a Compact Mapping Strategy.
Official implementation of "S²M²: Scalable Stereo Matching Model for Reliable Depth Estimation, ICCV 2025"
D-FINE-seg Object Detection and Segmentation Framework (Train, Export, Inference)
GIM: Learning Generalizable Image Matcher From Internet Videos (ICLR 2024 Spotlight)
Python wrapper for the NVIDIA cuSFM library
EfficientSAM3 compresses SAM3 into lightweight, edge-friendly models via progressive knowledge distillation for fast promptable concept segmentation and tracking.
High-performance multi-platform compiler for physics simulation
se7oluti0n / glim_localization
Forked from koide3/glimGLIM: with relocalization and wheel odometry
The official PyTorch implementation for the paper "Image-adaptive 3D Lookup Tables for Real-time Image Enhancement with Bilateral Grids" (accepted by ECCV 2024)
Image denoising using the Natural Image Noise Dataset
A fast and high quality GPU BVH builder implementing H-PLOC
Making it easier to work with slang in python
Sample benchmark demonstrating the VK_NV_cooperative_vector extension
Single-header dependency-free BVH construction and traversal library.
A set of GPU optimized space-filling-curve algorithms (e.g., Hilbert, Z-order) on PyTorch and OpenAI Triton.
[ICLR 2025]Source Code for the Paper "MIRReS: Inverse Rendering for Shape, Light, and Material Decomposition using Multi-Bounce Path Tracing and Reservoir Sampling"
a python BVH tree builder using slangpy to call cuda kernels, no optix build problems.