• Publications
  • Influence
Weakly Supervised Action Labeling in Videos under Ordering Constraints
TLDR
It is shown that the action label assignment can be determined together with learning a classifier for each action in a discriminative manner and evaluated on a new and challenging dataset of 937 video clips. Expand
Large-Margin Metric Learning for Constrained Partitioning Problems
TLDR
The experiments show how learning the metric can significantly improve performance on bioinformatics, video or image segmentation problems, and cast the metric learning problem as a large-margin structured prediction problem, with proper definition of regularizers and losses, leading to a convex optimization problem which can be solved efficiently. Expand
Weakly-Supervised Alignment of Video with Text
TLDR
This paper proposes a method for aligning the two modalities of video and text, i.e., automatically providing a time (frame) stamp for every sentence, and formulate this problem as an integer quadratic program, and solve its continuous convex relaxation using an efficient conditional gradient algorithm. Expand
Instance-Level Video Segmentation from Object Tracks
TLDR
This work addresses the problem of segmenting multiple object instances in complex videos by proposing a convex relaxation of this problem and solving it efficiently using the Frank-Wolfe algorithm. Expand
Metric Learning for Temporal Sequence Alignment
TLDR
This paper proposes to learn a Mahalanobis distance to perform alignment of multivariate time series, and proposes to use this metric learning framework to perform feature selection and, from basic audio features, build a combination of these with better alignment performance. Expand
A weakly-supervised discriminative model for audio-to-score alignment
TLDR
A new discriminative approach to the problem of audio-to-score alignment by extending the basic dynamic time warping algorithm to a convex problem that learns optimal classifiers for all events while jointly aligning files, using only weak supervision. Expand
Large-Margin Metric Learning for Partitioning Problems
TLDR
This paper focuses on partitioning problems based explicitly or implicitly on the minimization of Euclidean distortions, which include mean-based change-point detection, K-means, spectral clustering and normalized cuts, and learns a Mahalanobis metric for these unsupervised problems, leading to feature weighting and/or selection. Expand
Semidefinite and Spectral Relaxations for Multi-Label Classification
TLDR
This paper considers linear classifiers and proposes to learn a prior over the space of labels to directly leverage the performance of such methods, which takes the form of a quadratic function of the labels and permits to encode both attractive and repulsive relations between labels. Expand
...
1
2
...