Sequential Max-Margin Event Detectors

By Dong Huang, Shitong Yao, Yi Wang, Fernando De la Torre. 2014

94 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Wanqing Li
2. Pichao Wang
3. G. Wang
4. Philip O. Ogunbona
5. Jun Liu
6. Chang-Fu Tang
7. Giovanni Maria Farinella
8. Antonino Furnari

In Computer Vision and Pattern Recognition

1. G. Wang
2. Wanqing Li
3. Pichao Wang
4. Jun Liu
5. Philip O. Ogunbona
6. Chang-Fu Tang
7. Alex ChiChung Kot
8. J. Lai

Citations in the corpus

• RGB-D-based action recognition datasets: A survey (2016, background): “Due to the fact that there were very few publicly available sequential action dataset which can be used in the development and evaluation of detection algorithms, the Multi-modal action detection (MAD) Dataset [34](”
• Space-Time Representation of People Based on 3D Skeletal Data: A Review (2016, methodology): “ camera were also released to the public, including the MSR Daily Activity 3D [121], MSR Action Pairs [113], Online RGBD Action (ORGBD) [107], UTKinect-Action [124], Florence 3D-Action [118], CMU-MAD [104], UTD-MHAD [103], G3D/G3Di [105], [119], SPHERE [108], ChaLearn [111], RGB-D Person Re-identific”
• Rolling-Unrolling LSTMs for Action Anticipation from First-Person Video (2020, methodology): “Huang et al. [49] introduced Sequential Max-Margin Event Detectors, a method which performs early action detection by sequentially discarding classes until one class is identified as the detected one.”
• What Would You Expect? Anticipating Egocentric Actions With Rolling-Unrolling LSTMs and Modality Attention (2019, background): “The problem has been widely investigated in the domain of third person vision [2, 4, 6, 9, 19, 23, 24, 39, 46].”
• Modeling transition patterns between events for temporal human action segmentation and classification (2015, background): “For instance, Figure 2 shows an example of neutral, onset, offset, and peak segments for the action event corresponding to “crossing arms on chest.””
• Online Action Detection (2016, unknown): “dicult problem, that has not been studied in a real-world setting and with realistic data before. There are four main challenges. First, only partial actions are available (as previously stressed in [7,8,27]). Second, the negative data is highly variable and should not give rise to many false positi”
• Real-Time RGB-D Activity Prediction by Soft Regression (2016, background): “For example, [12, 15] explored the application of max-margin learning in early event recognition and detection.”
• Joint Geometrical and Statistical Alignment for Visual Domain Adaptation (2017, methodology): “We conduct experiments on the USPS→MNIST, W→A (SURF descriptor with linear kernel), and MSR→MAD datasets for illustration, which are shown in Figure 3.”