Abstract:Object detection is critical in autonomous driving, and it is more practical yet challenging to localize objects of unknown categories: an endeavour known as Class-Agnostic Object Detection (CAOD). Existing studies on CAOD predominantly rely on ordinary cameras, but these frame-based sensors usually have high latency and limited dynamic range, leading to safety risks in real-world scenarios. In this study, we turn to a new modality enabled by the so-called event camera, featured by its sub-millisecond latency and high dynamic range, for robust CAOD. We propose Detecting Every Object in Events (DEOE), an approach tailored for achieving high-speed, class-agnostic open-world object detection in event-based vision. Built upon the fast event-based backbone: recurrent vision transformer, we jointly consider the spatial and temporal consistencies to identify potential objects. The discovered potential objects are assimilated as soft positive samples to avoid being suppressed as background. Moreover, we introduce a disentangled objectness head to separate the foreground-background classification and novel object discovery tasks, enhancing the model's generalization in localizing novel objects while maintaining a strong ability to filter out the background. Extensive experiments confirm the superiority of our proposed DEOE in comparison with three strong baseline methods that integrate the state-of-the-art event-based object detector with advancements in RGB-based CAOD. Our code is available at https://github.com/Hatins/DEOE.
Abstract:The event camera produces a large dynamic range event stream with a very high temporal resolution discarding redundant visual information, thus bringing new possibilities for object detection tasks. However, the existing methods of applying the event camera to object detection tasks using deep learning methods still have many problems. First, existing methods cannot take into account objects with different velocities relative to the motion of the event camera due to the global synchronized time window and temporal resolution. Second, most of the existing methods rely on large parameter neural networks, which implies a large computational burden and low inference speed, thus contrary to the high temporal resolution of the event stream. In our work, we design a high-speed lightweight detector called Agile Event Detector (AED) with a simple but effective data augmentation method. Also, we propose an event stream representation tensor called Temporal Active Focus (TAF), which takes full advantage of the asynchronous generation of event stream data and is robust to the motion of moving objects. It can also be constructed without much time-consuming. We further propose a module called the Bifurcated Folding Module (BFM) to extract the rich temporal information in the TAF tensor at the input layer of the AED detector. We conduct our experiments on two typical real-scene event camera object detection datasets: the complete Prophesee GEN1 Automotive Detection Dataset and the Prophesee 1 MEGAPIXEL Automotive Detection Dataset with partial annotation. Experiments show that our method is competitive in terms of accuracy, speed, and the number of parameters simultaneously. Also by classifying the objects into multiple motion levels based on the optical flow density metric, we illustrated the robustness of our method for objects with different velocities relative to the camera.
Abstract:The basic Multi-Armed Bandits (MABs) problem is trying to maximize the rewards obtained from bandits with different unknown probability distributions of payoff for pulling different arms, given that only a finite number of attempts can be made. When studying trading algorithms in the market, we are looking at one of the most complex variants of MABs problems, namely the Non-stationary Continuum Bandits (NCBs) problem. The Bristol Stock Exchange (BSE) is a simple simulation of an electronic financial exchange based on a continuous double auction running via a limit order book. The market can be populated by automated trader agents with different trading algorithms. Within them, the PRSH algorithm embodies some basic ideas for solving NCBs problems. However, it faces the difficulty to adjust hyperparameters and adapt to changes in complex market conditions. We propose a new algorithm called PRB, which solves Continuum Bandits problem by Bayesian optimization, and solves Non-stationary Bandits problem by a novel "bandit-over-bandit" framework. With BSE, we use as many kinds of trader agents as possible to simulate the real market environment under two different market dynamics. We then examine the optimal hyperparameters of the PRSH algorithm and the PRB algorithm under different market dynamics respectively. Finally, by having trader agents using both algorithms trade in the market at the same time, we demonstrate that the PRB algorithm has better performance than the PRSH algorithm under both market dynamics. In particular, we perform rigorous hypothesis testing on all experimental results to ensure their correctness.