Multi-frame Motion Segmentation by Combining Two-Frame Results

Arrigoni, Federica; Ricci, Elisa; Pajdla, Tomáš

doi:10.1007/s11263-021-01544-x

Cited by 8 publications

(3 citation statements)

References 90 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many multi-shape/view/instance computer vision problems can be solved by synchronization, including shape (point set) alignment [36,43], structure from motion [11,8,39], multi-view matching [57,10], point cloud registration [37,46] and motion segmentation [44,2].…”

Section: Quantum Permutation Synchronization (Qps)mentioning

confidence: 99%

Q-FW: A Hybrid Classical-Quantum Frank-Wolfe for Quadratic Binary Optimization

Yurtsever¹,

Birdal²,

Golyanik³

2022

Preprint

View full text Add to dashboard Cite

We present a hybrid classical-quantum framework based on the Frank-Wolfe algorithm, Q-FW, for solving quadratic, linearly-constrained, binary optimization problems on quantum annealers (QA). The computational premise of quantum computers has cultivated the re-design of various existing vision problems into quantum-friendly forms. Experimental QA realisations can solve a particular non-convex problem known as the quadratic unconstrained binary optimization (QUBO). Yet a naive-QUBO cannot take into account the restrictions on the parameters. To introduce additional structure in the parameter space, researchers have crafted ad-hoc solutions incorporating (linear) constraints in the form of regularizers. However, this comes at the expense of a hyper-parameter, balancing the impact of regularization. To date, a true constrained solver of quadratic binary optimization (QBO) problems has lacked. Q-FW first reformulates constrained-QBO as a copositive program (CP), then employs Frank-Wolfe iterations to solve CP while satisfying linear (in)equality constraints. This procedure unrolls the original constrained-QBO into a set of unconstrained QUBOs all of which are solved, in a sequel, on a QA. We use D-Wave Advantage QA to conduct synthetic and real experiments on two important computer vision problems, graph matching and permutation synchronization, which demonstrate that our approach is effective in alleviating the need for an explicit regularization coefficient.

show abstract

Section: Quantum Permutation Synchronization (Qps)mentioning

confidence: 99%

Q-FW: A Hybrid Classical-Quantum Frank-Wolfe for Quadratic Binary Optimization

Yurtsever¹,

Birdal²,

Golyanik³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…However, they still require tracked trajectories. More recent approaches [64,65,66,67] avoid this requirement, for instance by using triplets of images [66]. However, despite achieving higher accuracy, trifocal tensors fail to classify a high percent of available pairs.…”

Section: List Of Acronymsmentioning

confidence: 99%

“…On the other hand, techniques based on combining two-view motion segmentation results to achieve multi-view motion segmentation [65,67] show that efficient two-view segmentation algorithms can help tackle challenges that multi-view segmentation methods struggle with.…”

Section: List Of Acronymsmentioning

confidence: 99%

Fast semi-algebraic clustering for efficient system identification and geometric scene understanding

Ozbay

View full text Add to dashboard Cite

I would like to begin by expressing my deepest gratitude to my advisor, Prof. Mario Sznaier. It was a privilege to work under his guidance throughout my PhD, learning not only how to conduct academic research but also how to tackle tough problems and brainstorm innovative solutions, including the occasional "Yet Another Stupid Idea" (YASI) when faced with unexpected challenges. I am particularly grateful for his encouragement to tackle tough problems and teaching me "Problems worthy of attack prove their worth by fighting back.". Additionally, I would like to thank Prof. Octavia Camps for her domain expertise in the computer vision field and her invaluable contributions to my research. It was a pleasure to work under her guidance and in collaboration with her team to learn about interesting problems in computer vision.I am thankful to my committee members, Prof. Taskin Padir and Prof. Rifat Sipahi, for their invaluable guidance and support. Prof. Padir's passion for robotics and his proactive initiatives was very inspiring. His robotics class provided me with extensive knowledge in the field. Additionally, collaborating with Prof. Sipahi in both academic and extracurricular endeavors has been enriching. I appreciate his support and his efforts in nurturing a sense of community among Turkish students at the university.My heartfelt appreciation goes to my family for their unwavering support and encouragement. My parents have been a constant source of strength, guiding me through challenges and celebrating my achievements. I am also grateful to my sister, who has been a faithful companion on this journey, offering a listening ear and unwavering support. It was a precious gift to have somebody by my side who went through the same process. I am grateful for their belief in me, which has been a driving force during difficult times. Special thanks to my sister and my brother-in-law for being like a second family to me in the US.Finally, I am deeply grateful to my fiancée, Ugurcan Ozdemir for literally being right by my side through all the worst and the best days since he entered my life. His support, constant presence, and unshakeable belief in me, especially during moments of doubt, have been invaluable. He has not only been a sounding board for navigating tough decisions but also provided technical support when needed. To my cherished friends who have become like family, especially

show abstract