Open-vocabulary semantic segmentation aims to segment an image into semantic regions according to text descriptions, which may not have been seen during training. Recent two-stage methods first generate class-agnostic mask proposals and then leverage pre-trained vision-language models, e.g., CLIP, to classify masked regions. We identify the performance bottleneck of this paradigm to be the pre-trained CLIP model, since it does not perform well on masked images. To address this, we propose to finetune CLIP on a collection of masked image regions and their corresponding text descriptions. We collect training data by mining an existing image-caption dataset (e.g., COCO Captions), using CLIP to match masked image regions to nouns in the image captions. Compared with the more precise and manually annotated segmentation labels with fixed classes (e.g., COCO-Stuff), we find our noisy but diverse dataset can better retain CLIP's generalization ability. Along with finetuning the entire model, we utilize the "blank" areas in masked images using a method we dub mask prompt tuning. Experiments demonstrate mask prompt tuning brings significant improvement without modifying any weights of CLIP, and it can further improve a fully finetuned model. In particular, when trained on COCO and evaluated on ADE20K-150, our best model achieves 29.6% mIoU, which is +8.5% higher than the previous state-of-the-art. For the first time, open-vocabulary generalist models match the performance of supervised specialist models in 2017 without dataset-specific adaptations.

Distributed learning on the edge often comprises self-centered devices (SCD) which learn local tasks independently and are unwilling to contribute to the performance of other SDCs. How do we achieve forward transfer at zero cost for the single SCDs? We formalize this problem as a Distributed Continual Learning scenario, where SCD adapt to local tasks and a CL model consolidates the knowledge from the resulting stream of models without looking at the SCD's private data. Unfortunately, current CL methods are not directly applicable to this scenario. We propose Data-Agnostic Consolidation (DAC), a novel double knowledge distillation method that consolidates the stream of SC models without using the original data. DAC performs distillation in the latent space via a novel Projected Latent Distillation loss. Experimental results show that DAC enables forward transfer between SCDs and reaches state-of-the-art accuracy on Split CIFAR100, CORe50 and Split TinyImageNet, both in reharsal-free and distributed CL scenarios. Somewhat surprisingly, even a single out-of-distribution image is sufficient as the only source of data during consolidation.

In recent years, car-sharing services have emerged as viable alternatives to private individual mobility, promising more sustainable and resource-efficient, but still comfortable transportation. Research on short-term prediction and optimization methods has improved operations and fleet control of car-sharing services; however, long-term projections and spatial analysis are sparse in the literature. We propose to analyze the average monthly demand in a station-based car-sharing service with spatially-aware learning algorithms that offer high predictive performance as well as interpretability. In particular, we compare the spatially-implicit Random Forest model with spatially-aware methods for predicting average monthly per-station demand. The study utilizes a rich set of socio-demographic, location-based (e.g., POIs), and car-sharing-specific features as input, extracted from a large proprietary car-sharing dataset and publicly available datasets. We show that the global Random Forest model with geo-coordinates as an input feature achieves the highest predictive performance with an R-squared score of 0.87, while local methods such as Geographically Weighted Regression perform almost on par and additionally yield exciting insights into the heterogeneous spatial distributions of factors influencing car-sharing behaviour. Additionally, our study offers effective as well as highly interpretable methods for diagnosing and planning the placement of car-sharing stations.

North Korea has tested a new underwater nuclear-capable attack drone designed to unleash a "radioactive tsunami" that would destroy enemy naval vessels and ports, state media has reported. During a military exercise conducted this week under the guidance of the country's leader Kim Jong Un, North Korea's military deployed and test-fired the new weapons system, the mission of which was to test the ability to set off a "super-scale" destructive blast and wave, the country's state news agency KCNA said on Friday. "This nuclear underwater attack drone can be deployed at any coast and port or towed by a surface ship for operation," KCNA said. The news agency said that during the exercise, the drone was put in the water off South Hamgyong province on Tuesday and cruised underwater for 59 hours and 12 minutes, at a depth of some 80 to 150 metres (260 to 490 feet), before detonating in waters off its east coast on Thursday. KCNA did not elaborate on the drone's nuclear capabilities.

Crippled by war and sanctions, Russia now faces evidence that its already-struggling space program is falling apart. In the past three months alone, Roscosmos has scrambled to resolve two alarming incidents. First, one of its formerly dependable Soyuz spacecraft sprang a coolant leak. Then the same thing happened on one of its Progress cargo ships. The civil space program's Soviet predecessor launched the first person into orbit, but with the International Space Station (ISS) nearing the end of its life, Russia's space agency is staring into the abyss.

Autonomous Underwater Vehicles (AUVs) conduct regular visual surveys of marine environments to characterise and monitor the composition and diversity of the benthos. The use of machine learning classifiers for this task is limited by the low numbers of annotations available and the many fine-grained classes involved. In addition to these challenges, there are domain shifts between image sets acquired during different AUV surveys due to changes in camera systems, imaging altitude, illumination and water column properties leading to a drop in classification performance for images from a different survey where some or all these elements may have changed. This paper proposes a framework to improve the performance of a benthic morphospecies classifier when used to classify images from a different survey compared to the training data. We adapt the SymmNet state-of-the-art Unsupervised Domain Adaptation method with an efficient bilinear pooling layer and image scaling to normalise spatial resolution, and show improved classification accuracy. We test our approach on two datasets with images from AUV surveys with different imaging payloads and locations. The results show that generic domain adaptation can be enhanced to produce a significant increase in accuracy for images from an AUV survey that differs from the training images.

Fox News correspondent Mike Tobin has the latest on tensions amid the Russia-Ukraine war on'Special Report.' The Russians planned the Black Sea drone attack carefully, probably for weeks. And watch out, China could try it next. As the admiral played by the late Sen. Fred D. Thompson said to Alec Baldwin's character in the classic movie "The Hunt for Red October," "The Russians don't do anything without a plan." Somebody on the Russian side thought this through.

Convolutional neural networks (CNNs) are a representative class of deep learning algorithms including convolutional computation that perform translation-invariant classification of input data based on their hierarchical architecture. However, classical convolutional neural network learning methods use the steepest descent algorithm for training, and the learning performance is greatly influenced by the initial weight settings of the convolutional and fully connected layers, requiring re-tuning to achieve better performance under different model structures and data. Combining the strengths of the simulated annealing algorithm in global search, we propose applying it to the hyperparameter search process in order to increase the effectiveness of convolutional neural networks (CNNs). In this paper, we introduce SA-CNN neural networks for text classification tasks based on Text-CNN neural networks and implement the simulated annealing algorithm for hyperparameter search. Experiments demonstrate that we can achieve greater classification accuracy than earlier models with manual tuning, and the improvement in time and space for exploration relative to human tuning is substantial.

The global battle for AI dominance is underway, according to author Paul Scharre, a former Army Ranger and current VP and director of studies at the Center for New American Security -- a think tank specializing in national security issues. Scharre previously served as a strategic planner at the Office of the Secretary of Defense, working to establish policies on unmanned and autonomous systems and emerging weapons technologies, and established DOD policies on intelligence, surveillance, and reconnaissance programs. In his latest book, "Four Battlegrounds: Power in the Age of Artificial Intelligence," Scharre explores how the international battle for the most powerful AI technology is changing global power dynamics. That battle, he says, is a global competition to seek the best and most efficient data, computing hardware, human talent, and institutions adopting AI technology -- which will determine the next global superpower. In your new book, you argue there's a battle for global power going on in the form of a revolution brought about by artificial intelligence.

Current work on using visual analytics to determine causal relations among variables has mostly been based on the concept of counterfactuals. As such the derived static causal networks do not take into account the effect of time as an indicator. However, knowing the time delay of a causal relation can be crucial as it instructs how and when actions should be taken. Yet, similar to static causality, deriving causal relations from observational time-series data, as opposed to designed experiments, is not a straightforward process. It can greatly benefit from human insight to break ties and resolve errors. We hence propose a set of visual analytics methods that allow humans to participate in the discovery of causal relations associated with windows of time delay. Specifically, we leverage a well-established method, logic-based causality, to enable analysts to test the significance of potential causes and measure their influences toward a certain effect. Furthermore, since an effect can be a cause of other effects, we allow users to aggregate different temporal cause-effect relations found with our method into a visual flow diagram to enable the discovery of temporal causal networks. To demonstrate the effectiveness of our methods we constructed a prototype system named DOMINO and showcase it via a number of case studies using real-world datasets. Finally, we also used DOMINO to conduct several evaluations with human analysts from different science domains in order to gain feedback on the utility of our system in practical scenarios.