A Hierarchical Representation for Future Action Prediction

By Tian Lan, Tsung-Chuan Chen, Silvio Savarese. 2014

342 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Biomedical Engineering, 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

In Biomedical Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

RGB-D-based Human Motion Recognition with Deep Learning: A Survey (2017, methodology): “Le et al. (2011) presented an extension of the independent subspace analysis algorithm (Theis, 2007) to learn invariant spatio-temporal features from unlabeled video data. They scale up the original ISA to larger input data by employing two important ideas from convolutional neural networks: convolu”
An Uncertain Future: Forecasting from Static Images Using Variational Autoencoders (2016, methodology): “Finally, supervised learning has also been used to model and forecast labeled human-human [2,16], and human-object [17,18] interactions.”
Unsupervised Semantic Parsing of Video Collections (2015, background): “Also, the previous works on finding action primitives such as [42, 60, 19, 32, 31] are primarily limited to discovering atomic sub-actions, and therefore, fail to identify complex and high-level parts of a long video.”
Newtonian Image Understanding: Unfolding the Dynamics of Objects in Static Images (2015, background): “Action Recognition: Early prediction of activities has been discussed in [29, 27, 16, 23].”
Learning to Decompose and Disentangle Representations for Video Prediction (2018, background): “Visual prediction is a core problem in computer vision that has been studied in several contexts, including activity prediction and early recognition [20, 30], human pose and trajectory forecasting [1, 18], and future frame prediction [22, 31, 39, 44].”
Dense Optical Flow Prediction from a Static Image (2015, background): “Others have looked at more semantic forms of prediction: that is, predict the action class of what is going to happen next [5, 14].”
Rolling-Unrolling LSTMs for Action Anticipation from First-Person Video (2020, background): “Lan et al. [53] proposed a hierarchical representation to anticipate future human actions from a still image or a short video clip.”
Anticipating the future by watching unlabeled video (2015, methodology): “[21] because their method uses an annotated bounding box around people.”
Flow-Grounded Spatial-Temporal Video Prediction from Still Images (2018, background): “The macroscopic analysis of prediction based on the given frame(s) can be predicting what event is going to happen [9,10,11], trajectory paths [12], or recognizing the type of human activities [13,14].”
Progressive Teacher-Student Learning for Early Action Prediction (2019, methodology): “Following the existing setting in [16, 24, 21, 32, 25, 26], we assume that each training video contains complete action execution and uniformly partition it into N shorter segments.”
Zero-Shot Anticipation for Instructional Activities (2018, background): “Prior work in forecasting activities before making any observations have been limited to simple movement primitives [28], or personal interactions [31, 55].”
Global Regularizer and Temporal-Aware Cross-Entropy for Skeleton-Based Early Action Recognition (2018, background): “It can also be used to enhance human-robot interaction, allowing robots to quickly respond to humans and improve user experience [2].”
r2p2: A ReparameteRized Pushforward Policy for Diverse, Precise Generative Path Forecasting (2018, background): “In [17,24,35,36], future activities are predicted via classification-based approaches.”
Part-Activated Deep Reinforcement Learning for Action Prediction (2018, background): “[20] exploited templates at multiple levels of granularities in a hierarchical representation, which can capture and compare human movements at different context levels.”
Action Recognition by Hierarchical Mid-Level Action Elements (2015, background): “A large body of work focuses on associating the entire video clip with a single class label from a pre-defined set of action categories (e.g., “take food from fridge” versus “cook food”) (Fig.”
Real-Time RGB-D Activity Prediction by Soft Regression (2016, methodology): “movements in a hierarchical manner and employ a max-margin learning framework to select the most discriminative features for prediction [3].”
Action Recognition with Image Based CNN Features (2015, background): “This is beneﬁcial for describing complex human actions [18, 17, 9].”