Research map / Video / Video segmentation and tracking
Quantum Motion Segmentation
Federica Arrigoni, Willi Menapace, Marcel Seelbach Benkner, Elisa Ricci, Vladislav Golyanik
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use sparse QUBO with known points per motion (v2) instead of dense v1 (ablation) | QuMoSeg · Q-MSEG (200 qubits) | accuracy: 0.99 -> 0.79 | Hurt |
| use sparse QUBO with known points per motion (v2) instead of dense v1 (ablation) | QuMoSeg · Q-MSEG (216 qubits) | accuracy: 0.64 -> 0.59 | Hurt |
| use sparse QUBO with known points per motion (v2) instead of dense v1 (ablation) | QuMoSeg · Hopkins sub-problems (180 qubits) | accuracy: 0.95 -> 0.88 | Hurt |
| use sparse QUBO v2 instead of dense v1 on smallest problem (ablation) | QuMoSeg · Q-MSEG (96 qubits) | accuracy: 0.97 -> 0.96 | No clear effect |
| solve the QUBO with simulated annealing instead of quantum annealing (v2) (ablation) | QuMoSeg-v2 · Q-MSEG (200 qubits) | accuracy: 0.79 -> 0.99 | Improved |
| solve the QUBO with simulated annealing instead of quantum annealing (v2) (ablation) | QuMoSeg-v2 · Q-MSEG (216 qubits) | accuracy: 0.59 -> 0.80 | Improved |
| solve the QUBO with simulated annealing instead of quantum annealing (v1) (ablation) | QuMoSeg-v1 · Q-MSEG (216 qubits) | accuracy: 0.64 -> 0.68 | Improved |
| solve the QUBO with simulated annealing instead of quantum annealing (v2) (ablation) | QuMoSeg-v2 · Hopkins sub-problems (180 qubits) | accuracy: 0.88 -> 1 | Improved |
| run on D-Wave Advantage4.1 instead of Advantage1.1 (ablation) | QuMoSeg-v2 · Q-MSEG (200 qubits) | accuracy: 0.71 -> 0.79 | Improved |
| run on D-Wave Advantage4.1 instead of Advantage1.1 (ablation) | QuMoSeg-v1 · Q-MSEG (96 qubits) | accuracy: 0.97 -> 0.97 | No clear effect |
| solve v2 binary QUBO without relaxation on AQC instead of relaxed Synch (ablation) | QuMoSeg-v2 vs Synch · Synthetic noise (96 qubits) | accuracy | Improved |
| increase problem size to 216 qubits (v1) | QuMoSeg-v1 · Q-MSEG | accuracy: 0.99 -> 0.64 | Hurt |
Closest papers by meaning
- An Iterative Quantum Approach for Transformation Estimation From Point Sets (CVPR 2022 · similarity 0.85)
- Probabilistic Sampling of Balanced K-Means using Adiabatic Quantum Computing (CVPR 2024 · similarity 0.80)
- QuCOOP: A Versatile Framework for Solving Composite and Binary-Parametrised Problems on Quantum Annealers (CVPR 2025 · similarity 0.80)
- Trajectory Inference with Smooth Schrödinger Bridges (ICML 2025 · similarity 0.79)
- Stochastic Quantum Sampling for Non-Logconcave Distributions and Estimating Partition Functions (ICML 2024 · similarity 0.79)
- Quantum Algorithms for Non-smooth Non-convex Optimization (NeurIPS 2024 · similarity 0.79)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
