Combined Person Classification with Airborne Optical Sectioning

Fully autonomous drones have been demonstrated to find lost or injured persons under strongly occluding forest canopy. Airborne Optical Sectioning (AOS), a novel synthetic aperture imaging technique, together with deep-learning-based classification enables high detection rates under realistic search-and-rescue conditions. We demonstrate that false detections can be significantly suppressed and true detections boosted by combining classifications from multiple AOS rather than single integral images. This improves classification rates especially in the presence of occlusion. To make this possible, we modified the AOS imaging process to support large overlaps between subsequent integrals, enabling real-time and on-board scanning and processing of groundspeeds up to 10 m/s.

STR-GODEs: Spatial-Temporal-Ridership Graph ODEs for Metro Ridership Prediction

The metro ridership prediction has always received extensive attention from governments and researchers. Recent works focus on designing complicated graph convolutional recurrent network architectures to capture spatial and temporal patterns. These works extract the information of spatial dimension well, but the limitation of temporal dimension still exists. We extended Neural ODE algorithms to the graph network and proposed the STR-GODEs network, which can effectively learn spatial, temporal, and ridership correlations without the limitation of dividing data into equal-sized intervals on the timeline. While learning the spatial relations and the temporal correlations, we modify the GODE-RNN cell to obtain the ridership feature and hidden states. Ridership information and its hidden states are added to the GODESolve to reduce the error accumulation caused by long time series in prediction. Extensive experiments on two large-scale datasets demonstrate the efficacy and robustness of our model.

Graph-guided Architecture Search for Real-time Semantic Segmentation

Designing a lightweight semantic segmentation network often requires researchers to find a trade-off between performance and speed, which is always empirical due to the limited interpretability of neural networks. In order to release researchers from these tedious mechanical trials, we propose a Graph-guided Architecture Search (GAS) pipeline to automatically search real-time semantic segmentation networks. Unlike previous works that use a simplified search space and stack a repeatable cell to form a network, we introduce a novel search mechanism with new search space where a lightweight model can be effectively explored through the cell-level diversity and latencyoriented constraint. Specifically, to produce the cell-level diversity, the cell-sharing constraint is eliminated through the cell-independent manner. Then a graph convolution network (GCN) is seamlessly integrated as a communication mechanism between cells. Finally, a latency-oriented constraint is endowed into the search process to balance the speed and performance. Extensive experiments on Cityscapes and CamVid datasets demonstrate that GAS achieves new state-of-the-art trade-off between accuracy and speed. In particular, on Cityscapes dataset, GAS achieves the new best performance of 73.3% mIoU with speed of 102 FPS on Titan Xp.

Energy-Based Open-World Uncertainty Modeling for Confidence Calibration

Confidence calibration is of great importance to the reliability of decisions made by machine learning systems. However, discriminative classifiers based on deep neural networks are often criticized for producing overconfident predictions that fail to reflect the true correctness likelihood of classification accuracy. We argue that such an inability to model uncertainty is mainly caused by the closed-world nature in softmax: a model trained by the cross-entropy loss will be forced to classify input into one of $K$ pre-defined categories with high probability. To address this problem, we for the first time propose a novel $K$+1-way softmax formulation, which incorporates the modeling of open-world uncertainty as the extra dimension. To unify the learning of the original $K$-way classification task and the extra dimension that models uncertainty, we propose a novel energy-based objective function, and moreover, theoretically prove that optimizing such an objective essentially forces the extra dimension to capture the marginal data distribution. Extensive experiments show that our approach, Energy-based Open-World Softmax (EOW-Softmax), is superior to existing state-of-the-art methods in improving confidence calibration.

Contrastive Learning for Sports Video: Unsupervised Player Classification

We address the problem of unsupervised classification of players in a team sport according to their team affiliation, when jersey colours and design are not known a priori. We adopt a contrastive learning approach in which an embedding network learns to maximize the distance between representations of players on different teams relative to players on the same team, in a purely unsupervised fashion, without any labelled data. We evaluate the approach using a new hockey dataset and find that it outperforms prior unsupervised approaches by a substantial margin, particularly for real-time application when only a small number of frames are available for unsupervised learning before team assignments must be made. Remarkably, we show that our contrastive method achieves 94% accuracy after unsupervised training on only a single frame, with accuracy rising to 97% within 500 frames (17 seconds of game time). We further demonstrate how accurate team classification allows accurate team-conditional heat maps of player positioning to be computed.

An overcome of far-distance limitation on tunnel CCTV-based accident detection in AI deep-learning frameworks

Tunnel CCTVs are installed to low height and long-distance interval. However, because of the limitation of installation height, severe perspective effect in distance occurs, and it is almost impossible to detect vehicles in far distance from the CCTV in the existing tunnel CCTV-based accident detection system (Pflugfelder 2005). To overcome the limitation, a vehicle object is detected through an object detection algorithm based on an inverse perspective transform by re-setting the region of interest (ROI). It can detect vehicles that are far away from the CCTV. To verify this process, this paper creates each dataset consisting of images and bounding boxes based on the original and warped images of the CCTV at the same time, and then compares performance of the deep learning object detection models trained with the two datasets. As a result, the model that trained the warped image was able to detect vehicle objects more accurately at the position far from the CCTV compared to the model that trained the original image.

Improving the Transient Times for Distributed Stochastic Gradient Methods

We consider the distributed optimization problem where $n$ agents each possessing a local cost function, collaboratively minimize the average of the $n$ cost functions over a connected network. Assuming stochastic gradient information is available, we study a distributed stochastic gradient algorithm, called exact diffusion with adaptive stepsizes (EDAS) adapted from the Exact Diffusion method and NIDS and perform a non-asymptotic convergence analysis. We not only show that EDAS asymptotically achieves the same network independent convergence rate as centralized stochastic gradient descent (SGD) for minimizing strongly convex and smooth objective functions, but also characterize the transient time needed for the algorithm to approach the asymptotic convergence rate, which behaves as $K_T=\mathcal{O}\left(\frac{n}{1-\lambda_2}\right)$, where $1-\lambda_2$ stands for the spectral gap of the mixing matrix. To the best of our knowledge, EDAS achieves the shortest transient time when the average of the $n$ cost functions is strongly convex and each cost function is smooth. Numerical simulations further corroborate and strengthen the obtained theoretical results.

Only Time Can Tell: Discovering Temporal Data for Temporal Modeling

Understanding temporal information and how the visual world changes over time is a fundamental ability of intelligent systems. In video understanding, temporal information is at the core of many current challenges, including compression, efficient inference, motion estimation or summarization. However, in current video datasets it has been observed that action classes can often be recognized without any temporal information from a single frame of video. As a result, both benchmarking and training in these datasets may give an unintentional advantage to models with strong image understanding capabilities, as opposed to those with strong temporal understanding. In this paper we address this problem head on by identifying action classes where temporal information is actually necessary to recognize them and call these "temporal classes". Selecting temporal classes using a computational method would bias the process. Instead, we propose a methodology based on a simple and effective human annotation experiment. We remove just the temporal information by shuffling frames in time and measure if the action can still be recognized. Classes that cannot be recognized when frames are not in order are included in the temporal Dataset. We observe that this set is statistically different from other static classes, and that performance in it correlates with a network's ability to capture temporal information. Thus we use it as a benchmark on current popular networks, which reveals a series of interesting facts. We also explore the effect of training on the temporal dataset, and observe that this leads to better generalization in unseen classes, demonstrating the need for more temporal data. We hope that the proposed dataset of temporal categories will help guide future research in temporal modeling for better video understanding.

Learning Sparse Interaction Graphs of Partially Observed Pedestrians for Trajectory Prediction

Multi-pedestrian trajectory prediction is an indispensable safety element of autonomous systems that interact with crowds in unstructured environments. Many recent efforts have developed trajectory prediction algorithms with focus on understanding social norms behind pedestrian motions. Yet we observe these works usually hold two assumptions that prevent them from being smoothly applied to robot applications: positions of all pedestrians are consistently tracked; the target agent pays attention to all pedestrians in the scene. The first assumption leads to biased interaction modeling with incomplete pedestrian data, and the second assumption introduces unnecessary disturbances and leads to the freezing robot problem. Thus, we propose Gumbel Social Transformer, in which an Edge Gumbel Selector samples a sparse interaction graph of partially observed pedestrians at each time step. A Node Transformer Encoder and a Masked LSTM encode the pedestrian features with the sampled sparse graphs to predict trajectories. We demonstrate that our model overcomes the potential problems caused by the assumptions, and our approach outperforms the related works in benchmark evaluation.

Using Quantum Mechanics to Cluster Time Series

In this article we present a method by which we can reduce a time series into a single point in $\mathbb{R}^{13}$. We have chosen 13 dimensions so as to prevent too many points from being labeled as "noise." When using a Euclidean (or Mahalanobis) metric, a simple clustering algorithm will with near certainty label the majority of points as "noise." On pure physical considerations, this is not possible. Included in our 13 dimensions are four parameters which describe the coefficients of a cubic polynomial attached to a Gaussian picking up a general trend, four parameters picking up periodicity in a time series, two each for amplitude of a wave and period of a wave, and the final five report the "leftover" noise of the detrended and aperiodic time series. Of the final five parameters, four are the centralized probabilistic moments, and the final for the relative size of the series. The first main contribution of this work is to apply a theorem of quantum mechanics about the completeness of the solutions to the quantum harmonic oscillator on $L^2(\mathbb{R})$ to estimating trends in time series. The second main contribution is the method of fitting parameters. After many numerical trials, we realized that methods such a Newton-Rhaphson and Levenberg-Marquardt converge extremely fast if the initial guess is good. Thus we guessed many initial points in our parameter space and computed only a few iterations, a technique common in Keogh's work on time series clustering. Finally, we have produced a model which gives incredibly accurate results quickly. We ackowledge that there are faster methods as well of more accurate methods, but this work shows that we can still increase computation speed with little, if any, cost to accuracy in the sense of data clustering.