In the paper, we aim to train a unified model that performs three tasks: Facial Action Units (FAU) prediction, seven basic facial expressions prediction, as well as valence and arousal prediction. The main challenge of this task is the lack of fully-annotated dataset. Most of existing datasets only contain one or two types of labels. To tackle this challenge, we propose an algorithm for the multitask model to learn from partial labels. The algorithm has two steps: first, we train a teacher model to perform all three tasks, where each instance is trained by the ground truth label of its corresponding task. Second, we refer to the outputs of the teacher model as the soft labels. We use the soft labels and the ground truths to train the student model. We find that the student model outperforms the teacher model on all the tasks, possibly due to the exposure to the full set of labels. Finally, we use ensemble modeling to boost the performance further on the three tasks.
Appearance-based gaze estimation from RGB images provides relatively unconstrained gaze tracking from commonly available hardware. The accuracy of subject-independent models is limited partly by small intra-subject and large inter-subject variations in appearance, and partly by a latent subject-dependent bias. To improve estimation accuracy, we propose to use dilated-convolutions in a deep convolutional neural network to capture subtle changes in the eye images, and a novel gaze decomposition method that decomposes the gaze angle into the sum of a subject-independent gaze estimate from the image and a subject-dependent bias. To further reduce estimation error, we propose a calibration method that estimates the bias from a few images taken as the subject gazes at only a few or even just a single gaze target. This significantly redues calibration time and complexity. Experiments on four datasets, including a new dataset we collected containing large variations in head pose and face location, indicate that even without calibration the estimator already outperforms state-of-the-art methods by more than 6.3%. The proposed calibration method is robust to the location of calibration target and reduces estimation error significantly (up to 35.6%), achieving state-of-the-art performance with much less calibration data than required by previously proposed methods.
Safe and efficient crowd navigation for mobile robot is a crucial yet challenging task. Previous work has shown the power of deep reinforcement learning frameworks to train efficient policies. However, their performance deteriorates when the crowd size grows. We suggest that this can be addressed by enabling the network to identify and pay attention to the humans in the crowd that are most critical to navigation. We propose a novel network utilizing a graph representation to learn the policy. We first train a graph convolutional network based on human gaze data that accurately predicts human attention to different agents in the crowd. Then we incorporate the learned attention into a graph-based reinforcement learning architecture. The proposed attention mechanism enables the assignment of meaningful weightings to the neighbors of the robot, and has the additional benefit of interpretability. Experiments on real-world dense pedestrian datasets with various crowd sizes demonstrate that our model outperforms state-of-art methods by 18.4% in task accomplishment and by 16.4% in time efficiency.
Appearance-based gaze estimation provides relatively unconstrained gaze tracking. However, subject-indepen\-dent models achieve limited accuracy partly due to individual variations. To improve estimation, we propose a novel gaze decomposition method and a single gaze point calibration method, motivated by our finding that the inter-subject squared bias exceeds the intra-subject variance for a subject-independent estimator. We decompose the gaze angle into a subject-dependent bias term and a subject-independent difference term between the gaze angle and the bias. The difference term is estimated by a deep convolutional network. For calibration-free tracking, we set the subject-dependent bias term to zero. For single gaze point calibration, we estimate the bias from a few images taken as the subject gazes at a point. Experiments on three datasets indicate that as a calibration-free estimator, the proposed method outperforms the state-of-the-art methods that use single model by up to $10.0\%$. The proposed calibration method is robust and reduces estimation error significantly (up to $35.6\%$), achieving state-of-the-art performance for appearance-based eye trackers with calibration.
Imitation learning by behavioral cloning is a prevalent method which has achieved some success in vision-based autonomous driving. The basic idea behind behavioral cloning is to have the neural network learn from observing a human expert's behavior. Typically, a convolutional neural network learns to predict the steering commands from raw driver-view images by mimicking the behaviors of human drivers. However, there are other cues, e.g. gaze behavior, available from human drivers that have yet to be exploited. Previous researches have shown that novice human learners can benefit from observing experts' gaze patterns. We present here that deep neural networks can also profit from this. We propose a method, gaze-modulated dropout, for integrating this gaze information into a deep driving network implicitly rather than as an additional input. Our experimental results demonstrate that gaze-modulated dropout enhances the generalization capability of the network to unseen scenes. Prediction error in steering commands is reduced by 23.5% compared to uniform dropout. Running closed loop in the simulator, the gaze-modulated dropout net increased the average distance travelled between infractions by 58.5%. Consistent with these results, we also found the gaze-modulated dropout net to have lower model uncertainty.
Appearance-based gaze estimation has attracted more and more attention because of its wide range of applications. The use of deep convolutional neural networks has improved the accuracy significantly. In order to improve the estimation accuracy further, we focus on extracting better features from eye images. Relatively large changes in gaze angles may result in relatively small changes in eye appearance. We argue that current architectures for gaze estimation may not be able to capture such small changes, as they apply multiple pooling layers or other downsampling layers so that the spatial resolution of the high-level layers is reduced significantly. To evaluate whether the use of features extracted at high resolution can benefit gaze estimation, we adopt dilated-convolutions to extract high-level features without reducing spatial resolution. In cross-subject experiments on the Columbia Gaze dataset for eye contact detection and the MPIIGaze dataset for 3D gaze vector regression, the resulting Dilated-Nets achieve significant (up to 20.8%) gains when compared to similar networks without dilated-convolutions. Our proposed Dilated-Net achieves state-of-the-art results on both the Columbia Gaze and the MPIIGaze datasets.
Many semantic video analysis tasks can benefit from multiple, heterogenous signals. For example, in addition to the original RGB input sequences, sequences of optical flow are usually used to boost the performance of human action recognition in videos. To learn from these heterogenous input sources, existing methods reply on two-stream architectural designs that contain independent, parallel streams of Recurrent Neural Networks (RNNs). However, two-stream RNNs do not fully exploit the reciprocal information contained in the multiple signals, let alone exploit it in a recurrent manner. To this end, we propose in this paper a novel recurrent architecture, termed Coupled Recurrent Network (CRN), to deal with multiple input sources. In CRN, the parallel streams of RNNs are coupled together. Key design of CRN is a Recurrent Interpretation Block (RIB) that supports learning of reciprocal feature representations from multiple signals in a recurrent manner. Different from RNNs which stack the training loss at each time step or the last time step, we propose an effective and efficient training strategy for CRN. Experiments show the efficacy of the proposed CRN. In particular, we achieve the new state of the art on the benchmark datasets of human action recognition and multi-person pose estimation.
The integration of information across multiple modalities and across time is a promising way to enhance the emotion recognition performance of affective systems. Much previous work has focused on instantaneous emotion recognition. The 2018 One-Minute Gradual-Emotion Recognition (OMG-Emotion) challenge, which was held in conjunction with the IEEE World Congress on Computational Intelligence, encouraged participants to address long-term emotion recognition by integrating cues from multiple modalities, including facial expression, audio and language. Intuitively, a multi-modal inference network should be able to leverage information from each modality and their correlations to improve recognition over that achievable by a single modality network. We describe here a multi-modal neural architecture that integrates visual information over time using an LSTM, and combines it with utterance level audio and text cues to recognize human sentiment from multimodal clips. Our model outperforms the unimodal baseline, achieving the concordance correlation coefficients (CCC) of 0.400 on the arousal task, and 0.353 on the valence task.
Human actions captured in video sequences are three-dimensional signals characterizing visual appearance and motion dynamics. To learn action patterns, existing methods adopt Convolutional and/or Recurrent Neural Networks (CNNs and RNNs). CNN based methods are effective in learning spatial appearances, but are limited in modeling long-term motion dynamics. RNNs, especially Long Short-Term Memory (LSTM), are able to learn temporal motion dynamics. However, naively applying RNNs to video sequences in a convolutional manner implicitly assumes that motions in videos are stationary across different spatial locations. This assumption is valid for short-term motions but invalid when the duration of the motion is long. In this work, we propose Lattice-LSTM (L2STM), which extends LSTM by learning independent hidden state transitions of memory cells for individual spatial locations. This method effectively enhances the ability to model dynamics across time and addresses the non-stationary issue of long-term motion dynamics without significantly increasing the model complexity. Additionally, we introduce a novel multi-modal training procedure for training our network. Unlike traditional two-stream architectures which use RGB and optical flow information as input, our two-stream model leverages both modalities to jointly train both input gates and both forget gates in the network rather than treating the two streams as separate entities with no information about the other. We apply this end-to-end system to benchmark datasets (UCF-101 and HMDB-51) of human action recognition. Experiments show that on both datasets, our proposed method outperforms all existing ones that are based on LSTM and/or CNNs of similar model complexities.
Optokinetic nystagmus (OKN) is an involuntary eye movement responsible for stabilizing retinal images in the presence of relative motion between an observer and the environment. Fully understanding the development of optokinetic nystagmus requires a neurally plausible computational model that accounts for the neural development and the behavior. To date, work in this area has been limited. We propose a neurally plausible framework for the joint development of disparity and motion tuning in the visual cortex, the optokinetic and vergence eye movements. This framework models the joint emergence of both perception and behavior, and accounts for the importance of the development of normal vergence control and binocular vision in achieving normal monocular OKN (mOKN) behaviors. Because the model includes behavior, we can simulate the same perturbations as performed in past experiments, such as artificially induced strabismus. The proposed model agrees both qualitatively and quantitatively with a number of findings from the literature on both binocular vision as well as the optokinetic reflex. Finally, our model also makes quantitative predictions about the OKN behavior using the same methods used to characterize the OKN in the experimental literature.