Multispectral to Hyperspectral using Pretrained Foundational model Ruben Gonzalez* 1, Conrad M Albrecht 1, Nassim Ait Ali Braham 1, Devyani Lambhate* 2, Joao Lucas de Sousa Almeida 2, Paolo Fraccaro 2, Benedikt Blumenstiel 2, Thomas Brunschwiler 2, and Ranjini Bangalore 2 1 Remote Sensing Technology Institute, German Aerospace Center (DLR), Germany 2 IBM Research Labs, India, U.K., Zurich, Brazil February 28, 2025 Abstract Hyperspectral imaging provides detailed spectral information, offering significant potential for monitoring greenhouse gases like CH 4 and NO 2. However, its application is constrained by limited spatial coverage and infrequent revisit times. In contrast, multispectral imaging delivers broader spatial and temporal coverage but lacks the spectral granularity required for precise GHG detection. To address these challenges, this study proposes Spectral and Spatial-Spectral transformer models that reconstructs hyperspectral data from multispectral inputs. The models in this paper are pretrained on EnMAP and EMIT datasets and fine-tuned on spatio-temporally aligned (Sentinel-2, EnMAP) and (HLS-S30, EMIT) image pairs respectively. Our model has the potential to enhance atmospheric monitoring by combining the strengths of hyperspectral and multispectral imaging systems. 1 Introduction Satellite images are being used to create detailed maps of Earth's surface.

ProbPNN: Enhancing Deep Probabilistic Forecasting with Statistical Information Benedikt Heidrich, Kaleb Phipps, Oliver Neumann, Marian Turowski, Ralf Mikut, Veit Hagenmeyer We combine statistical methods and deep learning-based forecasting methods to enhance probabilistic forecasts. We evaluate ProbPNN empirically on more than 1000 time series from an Electricity and a Traffic data set. On these datasets, the proposed ProbPNN outperforms existing state-of-the-art methods. Abstract Probabilistic forecasts are essential for various downstream applications such as business development, traffic planning, and electrical grid balancing. Many of these probabilistic forecasts are performed on time series data that contain calendar-driven periodicities. However, existing probabilistic forecasting methods do not explicitly take these periodicities into account. Therefore, in the present paper, we introduce a deep learning-based method that considers these calendar-driven periodicities explicitly. The present paper, thus, has a twofold contribution: First, we apply statistical methods that use calendar-driven prior knowledge to create rolling statistics and combine them with neural networks to provide better probabilistic forecasts.

According to a WHO prediction, by 2020, roughly 20% of India will suffer from some mental illness and 450 million people currently suffer from a mental illness, worldwide. These numbers are a wake-up call that psychology as an issue and psychologists as a profession must be taken seriously. Such helping professions are often considered as human channels. Unlike manual workers whose job responsibilities are being taken over by machines and AI bots, psychiatrists and counselors see no threat to their professions with the advancements of machine learning and artificial intelligence. According to an influential survey of the future of employment by Carl Benedikt Frey and Micheal Osborne who are Oxford economists, the probability that psychology could be automated in the future is only 0.43%.

Seven years ago, headlines were sensational and alarmist after Oxford academics Carl Benedikt Frey and Michael Osborne estimated that 47% of American jobs were at high risk of automation. While we agree that the rise of automation and intelligent technologies such as robots, AI, and machine learning are radically reshaping work across the globe, the hype continues to cloud the discussion. Alarmists continue to say that half of all jobs will disappear; technologists can't wait for the robots to arrive; policymakers are nervous; and business leaders see opportunity everywhere. Automation will create real change in how we get things done. Business and government leaders at all levels must plan for the transformation of human work.

History suggests that the worst fears of machines making humans obsolete don't come true. In some quarters, AI and robotics are considered the Fourth Industrial Revolution, following other big transformations in the 18th, 19th and 20th centuries. And while displacement of jobs occurred in each wave of new technology, new jobs emerged to balance out some of the pain. Concern about technology-driven mass unemployment "has proven to be exaggerated" throughout history, University of Oxford academics Carl Benedikt Frey and Michael Osborne wrote in an influential 2013 paper. Rather, technological progress "has vastly shifted the composition of employment, from agriculture and the artisan shop, to manufacturing and clerking, to service and management occupations," they wrote.

In the age of artificial intelligence, predicting which jobs will fall to automation is as much about what machines can do as it is about what they can't. More than half of all jobs in America -- both blue and white-collar -- are resistant to automation, according to an acclaimed study published in 2013 by two Oxford University researchers. Co-author Carl Benedikt Frey, who directs Oxford's Technology and Employment program, broke down three areas where human intelligence still beats artificial intelligence: perception and manipulation, social intelligence; and creativity. Each type has what Frey calls a "bottleneck," which slows the pace at which certain workforces can be automated. The premise is simple: Technology won't replace human workers if it can't do the job.

An intriguing set of questions that is being explored by researchers across the globe and is being discussed and brainstormed in various organisations and think tanks is: "what is the future of work"; "how forthcoming AI and Automation revolution will impact on the nature and structure of work"; and "what would be the impact of these changes on the fabric of society from social, economic and political perspectives". In a 2013 study "The Future of Employment: How Susceptible are Jobs to Computerisation?" researchers Dr Carl Benedikt Frey and Dr Michael Osborne made an important observation: about 47% jobs in the US will be lost to automation. Dr Carl Frey is the co-director of programme on technology and employment at Oxford Martin School at Oxford University. His research focuses on "how advances in digital technology are reshaping the nature of work and jobs and what that might mean for the future". In 2016, he was named the 2nd most influential young opinion leader by the Swedish business magazine Veckans Affärer.

Amazon announced last week that it will spend $700 million to train about 100,000 workers in the US by 2025, helping them move into more highly skilled jobs. The New York Times observed that with this program Amazon is acknowledging that "advances in automation technology will handle many tasks now done by people." The number of jobs which AI and machines will displace in the future has been the subject of numerous studies and surveys and op-eds and policy papers since 2013, when a pair of Oxford academics, Carl Benedikt Frey and Michael Osborne, estimated that 47% of American jobs are at high risk of automation by the mid-2030s. McKinsey Global Institute: between 40 million and 160 million women worldwide may need to transition between occupations by 2030, often into higher-skilled roles. Clerical work, done by secretaries, schedulers and bookkeepers, is an area especially susceptible to automation, and 72% of those jobs in advanced economies are held by women.

Amazon announced last week that it will spend $700 million to train about 100,000 workers in the US by 2025, helping them move into more highly skilled jobs. The New York Times observed that with this program Amazon is acknowledging that "advances in automation technology will handle many tasks now done by people." The number of jobs which AI and machines will displace in the future has been the subject of numerous studies and surveys and op-eds and policy papers since 2013, when a pair of Oxford academics, Carl Benedikt Frey and Michael Osborne, estimated that 47% of American jobs are at high risk of automation by the mid-2030s. McKinsey Global Institute: between 40 million and 160 million women worldwide may need to transition between occupations by 2030, often into higher-skilled roles. Clerical work, done by secretaries, schedulers and bookkeepers, is an area especially susceptible to automation, and 72% of those jobs in advanced economies are held by women.

IT IS ONE of the most widely quoted statistics of recent years. No report or conference presentation on the future of work is complete without it. Think-tanks, consultancies, government agencies and news outlets have pointed to it as evidence of an imminent jobs apocalypse. The finding--that 47% of American jobs are at high risk of automation by the mid-2030s--comes from a paper published in 2013 by two Oxford academics, Carl Benedikt Frey and Michael Osborne. It has since been cited in more than 4,000 other academic articles.