Large-scale visual place recognition (VPR) is inherently challenging because not all visual cues in the image are beneficial to the task. In order to highlight the task-relevant visual cues in the feature embedding, the existing attention mechanisms are either based on artificial rules or trained in a thorough data-driven manner. To fill the gap between the two types, we propose a novel Semantic Reinforced Attention Learning Network (SRALNet), in which the inferred attention can benefit from both semantic priors and data-driven fine-tuning. The contribution lies in two-folds. (1) To suppress misleading local features, an interpretable local weighting scheme is proposed based on hierarchical feature distribution. (2) By exploiting the interpretability of the local weighting scheme, a semantic constrained initialization is proposed so that the local attention can be reinforced by semantic priors. Experiments demonstrate that our method outperforms state-of-the-art techniques on city-scale VPR benchmark datasets.

Global technology company OPPO announced that the OPPO Research Institute has officially released its first 6G white paper - "6G AI-Cube Intelligent Networking". As one of the global and MENA region's telecommunications industry's first in-depth reports on how artificial intelligence (AI) can empower 6G network architecture, the white paper proposes a more detailed vision for the design of next-generation communication networks. OPPO has established a pre-research team to conduct preliminary research on 6G service and technology requirements, key technologies, and system features. The global smartphone leader believes that 6G will reshape the way people interact with AI, as it is utilised to serve the public through a myriad of applications. In June 2021, UAE telecoms provider Etisalat announced plans for 6G – stating that the network is expected to be even faster and support applications such as augmented and virtual reality, as well as AI infrastructure.

Some protesters in Minnesota set a fire last year. But then the surveillance footage from that day "set off a nearly yearlong, international manhunt...involving multiple federal agencies and Mexican police. The pursuit also involved a facial recognition system made by a Chinese company that has been blacklisted by the U.S. government." The New York Times tells the story of the couple who was eventually arrested: Ms. Yousif gave birth while on the run, and was separated from her baby for four months by the authorities. To prosecutors, the pursuit of Mr. Felan, who was charged with arson, and Ms. Yousif, who was charged with helping him flee, was a routine response to a case of property destruction...

Perception algorithms in autonomous vehicles are vital for the vehicle to understand the semantics of its surroundings, including detection and tracking of objects in the environment. The outputs of these algorithms are in turn used for decision-making in safety-critical scenarios like collision avoidance, and automated emergency braking. Thus, it is crucial to monitor such perception systems at runtime. However, due to the high-level, complex representations of the outputs of perception systems, it is a challenge to test and verify these systems, especially at runtime. In this paper, we present a runtime monitoring tool, PerceMon that can monitor arbitrary specifications in Timed Quality Temporal Logic (TQTL) and its extensions with spatial operators. We integrate the tool with the CARLA autonomous vehicle simulation environment and the ROS middleware platform while monitoring properties on state-of-the-art object detection and tracking algorithms.

Most data is automatically collected and only ever "seen" by algorithms. Yet, data compressors preserve perceptual fidelity rather than just the information needed by algorithms performing downstream tasks. In this paper, we characterize the bit-rate required to ensure high performance on all predictive tasks that are invariant under a set of transformations, such as data augmentations. Based on our theory, we design unsupervised objectives for training neural compressors. Using these objectives, we train a generic image compressor that achieves substantial rate savings (more than $1000\times$ on ImageNet) compared to JPEG on 8 datasets, without decreasing downstream classification performance.

Every day, 800 women and 6,700 newborns die from complications related to pregnancy or childbirth. A well-trained midwife can prevent most of these maternal and newborn deaths. Data science models together with logs generated by users of online learning applications for midwives can help to improve their learning competencies. The goal is to use these rich behavioral data to push digital learning towards personalized content and to provide an adaptive learning journey. In this work, we evaluate various forecasting methods to determine the interest of future users on the different kind of contents available in the app, broken down by profession and region.

Biological synaptic plasticity exhibits nonlinearities that are not accounted for by classic Hebbian learning rules. Here, we introduce a simple family of generalized, nonlinear Hebbian learning rules. We study the computations implemented by their dynamics in the simple setting of a neuron receiving feedforward inputs. We show that these nonlinear Hebbian rules allow a neuron to learn tensor decompositions of its higher-order input correlations. The particular input correlation decomposed, and the form of the decomposition, depend on the location of nonlinearities in the plasticity rule. For simple, biologically motivated parameters, the neuron learns tensor eigenvectors of higher-order input correlations. We prove that each tensor eigenvector is an attractor and determine their basins of attraction. We calculate the volume of those basins, showing that the dominant eigenvector has the largest basin of attraction. We then study arbitrary learning rules, and find that any learning rule that admits a finite Taylor expansion into the neural input and output also has stable equilibria at tensor eigenvectors of its higher-order input correlations. Nonlinearities in synaptic plasticity thus allow a neuron to encode higher-order input correlations in a simple fashion.

Successful sequential recommendation systems rely on accurately capturing the user's short-term and long-term interest. Although Transformer-based models achieved state-of-the-art performance in the sequential recommendation task, they generally require quadratic memory and time complexity to the sequence length, making it difficult to extract the long-term interest of users. On the other hand, Multi-Layer Perceptrons (MLP)-based models, renowned for their linear memory and time complexity, have recently shown competitive results compared to Transformer in various tasks. Given the availability of a massive amount of the user's behavior history, the linear memory and time complexity of MLP-based models make them a promising alternative to explore in the sequential recommendation task. To this end, we adopted MLP-based models in sequential recommendation but consistently observed that MLP-based methods obtain lower performance than those of Transformer despite their computational benefits. From experiments, we observed that introducing explicit high-order interactions to MLP layers mitigates such performance gap. In response, we propose the Multi-Order Interaction (MOI) layer, which is capable of expressing an arbitrary order of interactions within the inputs while maintaining the memory and time complexity of the MLP layer. By replacing the MLP layer with the MOI layer, our model was able to achieve comparable performance with Transformer-based models while retaining the MLP-based models' computational benefits.

AMLD Africa focuses on the application of machine learning and artificial intelligence in innovation and sustainable development in African countries, making it a particularly popular event for academia, industry, and business. The event will be totally virtual and in English. The talks will be broadcasted live and the workshops will be given virtually.

Picture this: Colossal, gas-powered autonomous robots bulldoze across acres of homogeneous farmland under a blackened sky that reeks of pollution. The trees have all been chopped down and there are no animals in sight. Pesticides are sprayed in excess because humans no longer tend to the fields. The machines do their jobs--producing massive amounts of food to feed our growing population--but it's not without ecological cost. Or, envision another future: Smaller robots cultivate mosaic plots of many different crops, working around the trees, streams, and wildlife of the natural landscape.