Shaofei Wang

Learning Optimal Parameters for Multi-target Tracking with Contextual Interactions

We describe an end-to-end framework for learning parameters of min-cost flow multi-target tracking problem with quadratic trajectory interactions including suppression of overlapping tracks and contextual cues about co-occurrence of different objects. Our approach utilizes structured prediction with a tracking-specific loss function to learn the complete set of model parameters. In this learning framework, we evaluate two different approaches to finding an optimal set of tracks under a quadratic model objective, one based on an linear program (LP) relaxation and the other based on novel greedy variants of dynamic programming that handle pairwise interactions. We find the greedy algorithms achieve almost equivalent accuracy to the LP relaxation while being up to 10

Accelerating Dynamic Programs via Nested Benders Decomposition with Application to Multi-Person Pose Estimation

\times

Multi-Person Pose Estimation via Column Generation.

× faster than a commercial LP solver. We evaluate trained models on three challenging benchmarks. Surprisingly, we find that with proper parameter learning, our simple data association model without explicit appearance/motion reasoning is able to achieve comparable or better accuracy than many state-of-the-art methods that use far more complex motion features or appearance affinity metric learning.

Exploiting skeletal structure in computer vision annotation with Benders decomposition.

We describe an end-to-end framework for learning parameters of min-cost flow multitarget tracking problem with quadratic trajectory interactions including suppression of overlapping tracks and contextual cues about co-occurrence of different objects. Our approach utilizes structured prediction with a tracking-specific loss function to learn the complete set of model parameters. Under our learning framework, we evaluate two different approaches to finding an optimal set of tracks under quadratic model objective based on an LP relaxation and a novel greedy extension to dynamic programming that handles pairwise interactions. We find the greedy algorithm achieves almost equivalent accuracy to the LP relaxation while being 2-7x faster than a commercial solver. We evaluate trained models on the challenging MOT and KITTI benchmarks. Surprisingly, we find that with proper parameter learning, our simple data-association model without explicit appearance/motion reasoning is able to outperform many state-of-the-art methods that use far more complex motion features and affinity metric learning.

Efficient Pose and Cell Segmentation using Column Generation.

We present a novel approach to solve dynamic programs (DP), which are frequent in computer vision, on tree-structured graphs with exponential node state space. Typical DP approaches have to enumerate the joint state space of two adjacent nodes on every edge of the tree to compute the optimal messages. Here we propose an algorithm based on Nested Benders Decomposition (NBD) that iteratively lower-bounds the message on every edge and promises to be far more efficient. We apply our NBD algorithm along with a novel Minimum Weight Set Packing (MWSP) formulation to a multi-person pose estimation problem. While our algorithm is provably optimal at termination it operates in linear time for practical DP problems, gaining up to 500\({\times }\) speed up over traditional DP algorithm which have polynomial complexity.