Predicting the Next Action by Modeling the Abstract Goal

2021 IEEE International Conference on Image Processing (ICIP)

We consider the task of training a neural network to anticipate human actions in video. This task is challenging given the complexity of video data, the stochastic nature of the future, and the limited amount of annotated training data. In this paper, we propose a novel knowledge distillation framework that uses an action recognition network to supervise the training of an action anticipation network, guiding the latter to attend to the relevant information needed for correctly anticipating the future actions. This framework is possible thanks to a novel loss function to account for positional shifts of semantic concepts in a dynamic video. The knowledge distillation framework is a form of self-supervised learning, and it takes advantage of unlabeled data. Experimental results on JHMDB and EPIC-KITCHENS dataset show the effectiveness of our approach.

IEEE Transactions on Image Processing, 2020

Future human action forecasting from partial observations of activities is an important problem in many practical applications such as assistive robotics, video surveillance and security. We present a method to forecast actions for the unseen future of the video using a neural machine translation technique that uses encoder-decoder architecture. The input to this model is the observed RGB video, and the objective is to forecast the correct future symbolic action sequence. Unlike prior methods that make action predictions for some unseen percentage of video one for each frame, we predict the complete action sequence that is required to accomplish the activity. We coin this task action sequence forecasting. To cater for two types of uncertainty in the future predictions, we propose a novel loss function. We show a combination of optimal transport and future uncertainty losses help to improve results. We evaluate our model in three challenging video datasets (Charades, MPII cooking and Breakfast). We extend our action sequence forecasting model to perform weakly supervised action forecasting on two challenging datasets, the Breakfast and the 50Salads. Specifically, we propose a model to predict actions of future unseen frames without using frame level annotations during training. Using Fisher vector features, our supervised model outperforms the state-of-the-art action forecasting model by 0.83% and 7.09% on the Breakfast and the 50Salads datasets respectively. Our weakly supervised model is only 0.6% behind the most recent state-of-the-art supervised model and obtains comparable results to other published fully supervised methods, and sometimes even outperforms them on the Breakfast dataset. Most interestingly, our weakly supervised model outperforms prior models by 1.04% leveraging on proposed weakly supervised architecture, and effective use of attention mechanism and loss functions.

Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, 2019

The idea of multi-sensor data fusion is to combine the data coming from different sensors to provide more accurate and complementary information to solve a specific task. Our goal is to build a shared representation related to data coming from different domains, such as images, audio signal, heart rate, acceleration, etc., in order to anticipate daily activities of a user wearing multimodal sensors. To this aim, we consider the Stanford-ECM Dataset which contains syncronized data acquired with different sensors: video, acceleration and heart rate signals. The dataset is adapted to our action prediction task by identifying the transitions from the generic "Unknown" class to a specific "Activity". We discuss and compare a Siamese Network with the Multi Layer Perceptron and the 1D CNN where the input is an unknown observation and the output is the next activity to be observed. The feature representations obtained with the considered deep architecture are classified with SVM or KNN classifiers. Experimental results pointed out that prediction from multimodal data seems a feasible task, suggesting that multimodality improves both classification and prediction. Nevertheless, the task of reliably predicting next actions is still open and requires more investigations as well as the availability of multimodal dataset, specifically built for prediction purposes.

Neurocomputing, 2021

To interact with humans in collaborative environments, machines need to be able to predict (i.e., anticipate) future events, and execute actions in a timely manner. However, the observation of the human limb movements may not be sufficient to anticipate their actions unambiguously. In this work, we consider two additional sources of information (i.e., context) over time, gaze, movement and object information, and study how these additional contextual cues improve the action anticipation performance. We address action anticipation as a classification task, where the model takes the available information as the input and predicts the most likely action. We propose to use the uncertainty about each prediction as an online decision-making criterion for action anticipation. Uncertainty is modeled as a stochastic process applied to a time-based neural network architecture, which improves the conventional class-likelihood (i.e., deterministic) criterion. The main contributions of this pape...

In contrast to the widely studied problem of recognizing an action given a complete sequence, action anticipation aims to identify the action from only partially available videos. As such, it is therefore key to the success of computer vision applications requiring to react as early as possible , such as autonomous navigation. In this paper, we propose a new action anticipation method that achieves high prediction accuracy even in the presence of a very small percentage of a video sequence. To this end, we develop a multi-stage LSTM architecture that leverages context-aware and action-aware features, and introduce a novel loss function that encourages the model to predict the correct class as early as possible. Our experiments on standard benchmark datasets evidence the benefits of our approach; We outperform the state-of-the-art action anticipation methods for early prediction by a relative increase in accuracy of 22.0% on JHMDB-21, 14.0% on UT-Interaction and 49.9% on UCF-101.

Cornell University - arXiv, 2022

Although human action anticipation is a task which is inherently multi-modal, state-of-the-art methods on well known action anticipation datasets leverage this data by applying ensemble methods and averaging scores of unimodal anticipation networks. In this work we introduce transformer based modality fusion techniques, which unify multi-modal data at an early stage. Our Anticipative Feature Fusion Transformer (AFFT) proves to be superior to popular score fusion approaches and presents state-of-the-art results outperforming previous methods on EpicKitchens-100 and EGTEA Gaze+. Our model is easily extensible and allows for adding new modalities without architectural changes. Consequently, we extracted audio features on EpicKitchens-100 which we add to the set of commonly used features in the community. 1

British Machine Vision Conference (BMVC), 2019

We propose a novel neural memory network based framework for future action sequence forecasting. This is a challenging task where we have to consider short-term, within sequence relationships as well as relationships in between sequences, to understand how sequences of actions evolve over time. To capture these relationships effectively , we introduce neural memory networks to our modelling scheme. We show the significance of using two input streams, the observed frames and the corresponding action labels, which provide different information cues for our prediction task. Furthermore, through the proposed method we effectively map the long-term relationships among individual input sequences through separate memory modules, which enables better fusion of the salient features. Our method outperforms the state-of-the-art approaches by a large margin on two publicly available datasets: Breakfast and 50 Salads.

arXiv (Cornell University), 2022

We address the following action-effect prediction task. Given an image depicting an initial state of the world and an action expressed in text, predict an image depicting the state of the world following the action. The prediction should have the same scene context as the input image. We explore the use of the recently proposed GLIDE model for performing this task. GLIDE is a generative neural network that can synthesize (inpaint) masked areas of an image, conditioned on a short piece of text. Our idea is to mask-out a region of the input image where the effect of the action is expected to occur. GLIDE is then used to inpaint the masked region conditioned on the required action. In this way, the resulting image has the same background context as the input image, updated to show the effect of the action. We give qualitative results from experiments using the EPIC dataset of ego-centric videos labelled with actions.

Predicting another person's upcoming action to build an appropriate response is a regular occurrence in the domain of motor control. In this review we discuss conceptual and experimental approaches aiming at the neural basis of predicting and learning to predict upcoming movements by their observation.

Robots that interact with people must flexibly respond to requests by planning in stochastic state spaces that are often too large to solve for optimal behavior. In this work, we develop a framework for goal and state dependent action priors that can be used to prune away irrelevant actions based on the robot's current goal, thereby greatly accelerating planning in a variety of complex stochastic environments. Our framework allows these goal-based action priors to be specified by an expert or to be learned from prior experience in related problems. We evaluate our approach in the video game Minecraft, whose complexity makes it an effective robot simulator. We also evaluate our approach in a robot cooking domain that is executed on a two-handed manipulator robot. In both cases, goal-based action priors enhance baseline planners by dramatically reducing the time taken to find a near-optimal plan.

IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021

Human actions involving hand manipulations are structured according to the making and breaking of hand-object contact, and human visual understanding of action is reliant on anticipation of contact as is demonstrated by pioneering work in cognitive science. Taking inspiration from this, we introduce representations and models centered on contact, which we then use in action prediction and anticipation. We annotate a subset of the EPIC Kitchens dataset to include time-to-contact between hands and objects, as well as segmentations of hands and objects. Using these annotations we train the Anticipation Module, a module producing Contact Anticipation Maps and Next Active Object Segmentations-novel low-level representations providing temporal and spatial characteristics of anticipated near future action. On top of the Anticipation Module we apply Egocentric Object Manipulation Graphs (Ego-OMG), a framework for action anticipation and prediction. Ego-OMG models longer term temporal semantic relations through the use of a graph modeling transitions between contact delineated action states. Use of the Anticipation Module within Ego-OMG produces state-of-the-art results, achieving 1 st and 2 nd place on the unseen and seen test sets, respectively, of the EPIC Kitchens Action Anticipation Challenge, and achieving state-of-the-art results on the tasks of action anticipation and action prediction over EPIC Kitchens. We perform ablation studies over characteristics of the Anticipation Module to evaluate their utility.

2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2019

Anticipating actions before they are executed is crucial for a wide range of practical applications including autonomous driving and robotics. While most prior work in this area requires partial observation of executed actions, in the paper we focus on anticipating actions seconds before they start. Our proposed approach is the fusion of a purely anticipatory model with a complementary model constrained to reason about the present. In particular, the latter predicts present action and scene attributes, and reasons about how they evolve over time. By doing so, we aim at modeling action anticipation at a more conceptual level than directly predicting future actions. Our model outperforms previously reported methods on the EPIC-KITCHENS and Breakfast datasets.

arXiv (Cornell University), 2021

Humans typically perceive the establishment of an action in a video through the interaction between an actor and the surrounding environment. An action only starts when the main actor in the video begins to interact with the environment, while it ends when the main actor stops the interaction. Despite the great progress in temporal action proposal generation, most existing works ignore the aforementioned fact and leave their model learning to propose actions as a black-box. In this paper, we make an attempt to simulate that ability of a human by proposing Actor Environment Interaction (AEI) network to improve the video representation for temporal action proposals generation. AEI contains two modules, i.e., perception-based visual representation (PVR) and boundary-matching module (BMM). PVR represents each video snippet by taking human-human relations and humans-environment relations into consideration using the proposed adaptive attention mechanism. Then, the video representation is taken by BMM to generate action proposals. AEI is comprehensively evaluated in ActivityNet-1.3 and THUMOS-14 datasets, on temporal action proposal and detection tasks, with two boundary-matching architectures (i.e., CNN-based and GCN-based) and two classifiers (i.e., Unet and P-GCN). Our AEI robustly outperforms the state-of-the-art methods with remarkable performance and generalization for both temporal action proposal generation and temporal action detection. Source code is available at 1 .

IEEE transactions on pattern analysis and machine intelligence, 2018

We propose a novel approach for predicting on-going action with the assistance of a low-cost depth camera. Our approach introduces a soft regression-based early prediction framework. In this framework, we estimate soft labels for the subsequences at different progress levels, jointly learned with an action predictor. Our formulation of soft regression framework 1) overcomes a usual assumption in existing early action prediction systems that the progress level of on-going sequence is given in the testing stage; and 2) presents a theoretical framework to better resolve the ambiguity and uncertainty of subsequences at early performing stage. The proposed soft regression framework is further enhanced in order to take the relationships among subsequences and the discrepancy of soft labels over different classes into consideration, so that a Multiple Soft labels Recurrent Neural Network (MSRNN) is finally developed. For real-time performance, we also introduce "local accumulative fra...

Cornell University - arXiv, 2019

Task-oriented dialog presents a difficult challenge encompassing multiple problems including multi-turn language understanding and generation, knowledge retrieval and reasoning, and action prediction. Modern dialog systems typically begin by converting conversation history to a symbolic object referred to as belief state by using supervised learning. The belief state is then used to reason on an external knowledge source whose result along with the conversation history is used in action prediction and response generation tasks independently. Such a pipeline of individually optimized components not only makes the development process cumbersome but also makes it non-trivial to leverage session-level user reinforcement signals. In this paper, we develop Neural Assistant: a single neural network model that takes conversation history and an external knowledge source as input and jointly produces both text response and action to be taken by the system as output. The model learns to reason on the provided knowledge source with weak supervision signal coming from the text generation and the action prediction tasks, hence removing the need for belief state annotations. In the MultiWOZ dataset, we study the effect of distant supervision, and the size of knowledge base on model performance. We find that the Neural Assistant without belief states is able to incorporate external knowledge information achieving higher factual accuracy scores compared to Transformer. In settings comparable to reported baseline systems, Neural Assistant when provided with oracle belief state significantly improves language generation performance. * Equal contribution † Work done when all authors were at Google 3 We ignore speech-to-text and text-to-speech components in this work.