Forecasting Events Using an Augmented Hidden Conditional Random Field

@inproceedings{Wei2014ForecastingEU,
  title={Forecasting Events Using an Augmented Hidden Conditional Random Field},
  author={Xinyu Wei and Patrick Lucey and Stephen Vidas and Stuart Morgan and Sridha Sridharan},
  booktitle={ACCV},
  year={2014}
}
In highly dynamic and adversarial domains such as sports, short-term predictions are made by incorporating both local immediate as well global situational information. For forecasting complex events, higher-order models such as Hidden Conditional Random Field (HCRF) have been used to good effect as capture the long-term, high-level semantics of the signal. However, as the prediction is based solely on the hidden layer, fine-grained local information is not incorporated which reduces its… 
Joint Prediction of Activity Labels and Starting Times in Untrimmed Videos
TLDR
Experiments demonstrate that the framework for joint prediction of activity label and starting time improves the performance of both, and outperforms the state-of-the-arts.
Learning to Predict Human Behavior from Video
TLDR
This thesis presents a framework for learning a representation of human dynamics that can be used to estimate the 3d pose and shape of people moving in videos, and use to hallucinate the motion surrounding a single-frame snapshot.
Forecasting the Next Shot Location in Tennis Using Fine-Grained Spatiotemporal Tracking Data
TLDR
This research uses ball and player tracking data from “Hawk-Eye” to discover unique player styles and predict within-point events, and moves beyond current analysis that only incorporates coarse match statistics and uses spatial and temporal information which better characterizes the tactics and tendencies of each player.
A poisson process model for activity forecasting
TLDR
This paper proposes an approach for predicting the starting time of the next unobserved activity without assuming that the authors know its label, and model activities occurring at a variable rate using a Log-Gaussian Cox Process and learn the rate function from the training data.
Prediction and Description of Near-Future Activities in Video
Captioning Near-Future Activity Sequences
TLDR
This work proposes a system that can infer the labels and the captions of a sequence of future activities and demonstrates that both the label prediction framework and captioning framework outperforms the state-of-the-arts.
When will you do what? - Anticipating Temporal Occurrences of Activities
TLDR
This paper proposes two methods to predict a considerably large amount of future actions and their durations using a CNN and an RNN that are trained to learn future video labels based on previously seen content.
Action anticipation using latent goal learning
TLDR
An action anticipation model that uses goal information in an effective manner that is able to predict the next noun better than existing approaches on both seen and unseen kitchens in the test set.
Leveraging the Present to Anticipate the Future in Videos
TLDR
This work focuses on anticipating actions seconds before they start, and proposes a fusion of a purely anticipatory model with a complementary model constrained to reason about the present that predicts present action and scene attributes, and reasons about how they evolve over time.
Fine-Grained Retrieval of Sports Plays using Tree-Based Alignment of Trajectories
TLDR
The results show that the proposed tree-based method uses a hierarchical structure to perform multi-agent data alignment and partitioning in a coarse-to-fine fashion boosts performance in retrieving similar sports plays -- especially in interactive situations where the user selects a subset of trajectories compared to current state-of-the-art methods.
...
...

References

SHOWING 1-10 OF 21 REFERENCES
Predicting Shot Locations in Tennis Using Spatiotemporal Data
TLDR
A method which predicts the location of a future shot based on the spatiotemporal parameters of the incoming shots (i.e. shot speed, location, angle and feet location) from such a vision system is presented.
Activity Forecasting
TLDR
The unified model uses state-of-the-art semantic scene understanding combined with ideas from optimal control theory to achieve accurate activity forecasting and shows how the same techniques can improve the results of tracking algorithms by leveraging information about likely goals and trajectories.
Human activity prediction: Early recognition of ongoing activities from streaming videos
  • M. Ryoo
  • Computer Science
    2011 International Conference on Computer Vision
  • 2011
TLDR
The new recognition methodology named dynamic bag-of-words is developed, which considers sequential nature of human activities while maintaining advantages of the bag- of-words to handle noisy observations, and reliably recognizes ongoing activities from streaming videos with a high accuracy.
“ Sweet-Spot ” : Using Spatiotemporal Data to Discover and Predict Shots in Tennis
TLDR
This paper uses ball and player tracking data from “Hawk-Eye” to discover unique player styles and predict within-point events using a probabilistic graphical model and uses model adaptation to enhance the prediction.
You'll never walk alone: Modeling social behavior for multi-target tracking
TLDR
A model of dynamic social behavior, inspired by models developed for crowd simulation, is introduced, trained with videos recorded from birds-eye view at busy locations, and applied as a motion model for multi-people tracking from a vehicle-mounted camera.
Hidden Conditional Random Fields for Gesture Recognition
TLDR
This paper derives a discriminative sequence model with a hidden state structure, and demonstrates its utility both in a detection and in a multi-way classification formulation.
Motion fields to predict play evolution in dynamic sport scenes
TLDR
This work proposes a novel approach to detect the locations of where the play evolution will proceed, e.g. where interesting events will occur, by tracking player positions and movements over time, and generates a dense motion field.
Hidden-state Conditional Random Fields
TLDR
This work presents a discriminative latent variable model for classification problems in structured domains where inputs can be represented by a graph of local observations and evaluates the model on object detection and gesture recognition tasks.
Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data
TLDR
This work presents iterative parameter estimation algorithms for conditional random fields and compares the performance of the resulting models to HMMs and MEMMs on synthetic and natural-language data.
Large-Scale Analysis of Formations in Soccer
TLDR
This work finds the most likely formation patterns of a team associated with match events across nearly 14 hours of continuous player and ball tracking data in soccer and shows that it can accurately segment a match into distinct game phases and detect highlights.
...
...