Most models tasked to ground referential utterances in 2D and 3D scenes learn to select the referred object from a pool of object proposals provided by a pre-trained detector. This is limiting because an utterance may refer to visual entities at various levels of granularity, such as the chair, the leg of the chair, or the tip of the front leg of the chair, which may be missed by the detector. We propose a language grounding model that attends on the referential utterance and on the object proposal pool computed from a pre-trained detector to decode referenced objects with a detection head, without selecting them from the pool. In this way, it is helped by powerful pre-trained object detectors without being restricted by their misses. We call our model Bottom Up Top Down DEtection TRansformers (BUTD-DETR) because it uses both language guidance (top down) and objectness guidance (bottom-up) to ground referential utterances in images and point clouds. Moreover, BUTD-DETR casts object detection as referential grounding and uses object labels as language prompts to be grounded in the visual scene, augmenting supervision for the referential grounding task in this way. The proposed model sets a new state-of-the-art across popular 3D language grounding benchmarks with significant performance gains over previous 3D approaches (12.6% on SR3D, 11.6% on NR3D and 6.3% on ScanRefer). When applied in 2D images, it performs on par with the previous state of the art. We ablate the design choices of our model and quantify their contribution to performance. Our code and checkpoints can be found at the project website https://butd-detr.github.io.

Causal Machine Learning (CausalML) is an umbrella term for machine learning methods that formalize the data-generation process as a structural causal model (SCM). This perspective enables us to reason about the effects of changes to this process (interventions) and what would have happened in hindsight (counterfactuals). We categorize work in CausalML into five groups according to the problems they address: (1) causal supervised learning, (2) causal generative modeling, (3) causal explanations, (4) causal fairness, and (5) causal reinforcement learning. We systematically compare the methods in each category and point out open problems. Further, we review data-modality-specific applications in computer vision, natural language processing, and graph representation learning. Finally, we provide an overview of causal benchmarks and a critical discussion of the state of this nascent field, including recommendations for future work.

As part of ongoing efforts to diversify their economies and build a platform for sustainable future growth, MENA nations are increasingly turning towards artificial intelligence (AI). A slew of recent investment and initiatives – primarily in academia and the government, but also in the private sector – has reinvigorated interest from industry leaders around the globe in the potential for AI to strengthen the efficiency and sustainability of MENA economies. According to a report from the Economist Impact Unit (EIU) and Google published earlier this year, AI could bring about an additional $320bn in economic growth in the MENA region by 2030. Many long-term economic strategies in the region target high-value sectors with the potential to benefit from the Fourth Industrial Revolution – a raft of technological advancements in AI, data and cloud computing that merge the physical, digital and biological worlds. In recent years the UAE, Saudi Arabia, Qatar and Egypt have published ambitious, government-driven strategies to develop AI.

Check out what's clicking on Foxnews.com. Rep. Matt Gaetz, R-Fla., hammered his former Republican colleague-turned-Disney lobbyist in a tense exchange on Tuesday. Gaetz turned the heat up on former House Judiciary Committee chairman Bob Goodlatte, R-Va., during the committee's Tuesday hearing on government access to Americans' personal data. Goodlatte, who retired in 2018, was asked by Gaetz about his work with Disney and what the Mouse House is doing to prevent children's data from being sold. Gaetz said he was worried about "private entities" trying to "program" people and "make them think a certain way" by harmfully utilizing people's data.

We asked some of TechRepublic's contributing writers to offer their expertise on software, networking solutions and hardware. They weighed in on the usefulness and standout features of some of the best technology for remote work, cybersecurity, collaboration and DaaS; in addition, Matt Asay compiled a list of the most innovative artificial intelligence and machine learning companies. The following tech products and companies were deemed the best of its category. In Matt Asay's article about the eight AI and machine learning companies, he ranked Hugging Face as the most innovative. Here's what he had to say: "Hugging Face, which started as a chatbot and evolved to offer a registry of NLP models used to deliver those chatbots, is now on track to become the GitHub of ML. More than 10,000 companies use Hugging Face to privately collaborate on ML applications."

Research into Artificial Intelligence (AI) has been ongoing for decades, with early proposals dating back to 1950. However, only in recent years, it has seen a resurgence in popularity thanks to the increased availability of computing power and the growth of big data and machine learning. AI is the ability of machines to perform tasks that ordinarily require human intelligence, such as understanding natural language and recognizing objects. With the rapid expansion of AI, there are opportunities for businesses and individuals alike to capitalize on its capabilities. AI is a field of computer science and engineering focused on the creation of intelligent agents, which are systems that can reason, learn, and act autonomously.

Federal agencies that want to successfully scale and implement cloud computing systems into existing infrastructure can do so through several key practices, including designating expert teams, two-factor authentication, and enhanced education opportunities among users. Outlined in a White House report published earlier this month, officials documented how cloud computing systems can support further federal research and development in artificial intelligence, a goal within the broader Biden administration. Authored by the Machine Learning and Artificial Intelligence Subcommittee within the National Science and Technology Council, the report notes that leveraging cloud computing technology can enable better on-demand resources for researchers working with AI technologies. It went on to highlight opportunities for public agencies looking to bolster AI research efforts with advanced computing systems. "Agencies that have undertaken early efforts to leverage commercial cloud computing resources to advance AI R&D have commonly experienced benefits to their investments in terms of providing internal and external researchers persistent, on-demand access to cutting-edge capabilities, accelerating experimentation and the use of AI in new domains, and enabling reproducibility and scalability of the research activities and results," the report explains.

DALL-E2, the AI image tool, generated these images of a giraffe shopping in a grocery store. When the Silicon Valley research lab OpenAI unveiled DALL-E earlier this year, it wowed the internet. The tool is seen as one of the most advanced artificial intelligence systems for creating images in the world. Type a description, and DALL-E instantly produces professional-looking art or hyperrealistic photographs. "It's incredibly powerful," said Hany Farid, a digital forensics expert at the University of California, Berkeley.

Speaking at the Aspen Security Forum, Gen. John Raymond the head of U.S. Space Command, discussed the main issues in space and said there was a need for a rules-based order and described the present situation as being like the'wild, Wild West.' (Video courtesy: Aspen Security Forum.) Chief of Space Operations for the U.S. Space Force Gen. John'Jay' Raymond stressed the need for international norms when it comes to space operations, while pointing to problems posed by Russia and China. Addressing the Aspen Security Forum on Tuesday, Raymond said China was growing its program at a fast pace, explaining "China has gone from zero to 60 very quickly, and they are clearly our pacing challenge because…they're moving at speed they have the economy to support the development. "They're really doing two things: the first thing they're doing is they're building space capabilities for their own use, so just like we've enjoyed space capabilities that we've been able to integrate, China has built a space program to do the same thing," Raymond said while noting this "provides them advantage and that provides risk to our forces. The other thing that they're doing, they have seen the advantages that space has provided us as: we've integrated space and cyber and multi domain operations, and to be honest they don't like what they see." Raymond further explained that while space operations are hardly something new, the area has exploded in recent years to the point of being far more difficult to manage. "One of the challenges is there are no rules or very few rules," Raymond said. Raymond said that the U.S. is trying to lead the way, and that there have been discussions among other countries and the United Nations. "This is something that we're trying to establish the the the norms, if you will, the rules of the road," he said. One example Raymond discussed was the issue of space debris. He mentioned how Vice President Kamala Harris announced that the U.S. will not conduct destructive, direct-ascent anti-satellite (ASAT) missile testing while calling on other nations to make similar commitments. These tests create long-lasting debris in space that can threaten existing satellites and pose dangers to astronauts. Russia conducted such a test in 2021, and China did the same in 2007. Roger Towberman displays his insignia during a presentation of the United States Space Force flag in the Oval Office of the White House in Washington, D.C., May 15, 2020. Raymond said Russia's test resulted in blowing up a satellite into more than 1,500 pieces, while China's test created 3,000 pieces of debris. Raymond added that the U.S. has been trying to manage these sorts of situations. "We act as the space traffic control for the world.

The cyber industry has a long history of dealing with data and information in both paper and in digital formats. The information in our DNA is different. When compromised, it is not something we can change like a lost password or username or even a Social Security Number (SSN). We cannot even conceive of its value. There is no practical way of quantifying the financial impact of its loss.