Welcome to the second of our monthly digests, designed to keep you up-to-date with the happenings in the AI world. You can catch up with any AIhub stories you may have missed, get the low-down on recent conferences, and generally immerse yourself in all things AI. You may be aware that we are running a focus series on the UN sustainable development goals (SDG). Each month we tackle a different SDG and cover some of the AI research linked to that particular goal. In February it was the turn of climate action.

"Ordinary people here in China aren't happy about this technology but they have no choice. If the police say there have to be cameras in a community, people will just have to live with it. So says Chen Wei at Taigusys, a company specialising in emotion recognition technology, the latest evolution in the broader world of surveillance systems that play a part in nearly every aspect of Chinese society. Emotion-recognition technologies – in which facial expressions of anger, sadness, happiness and boredom, as well as other biometric data are tracked – are supposedly able to infer a person's feelings based on traits such as facial muscle movements, vocal tone, body movements and other biometric signals. It goes beyond facial-recognition technologies, which simply compare faces to determine a match. But similar to facial recognition, it involves the mass collection of sensitive personal data to track, monitor and profile people and uses machine learning to analyse expressions and other clues. The industry is booming in China, where since at least 2012, figures including President Xi Jinping have emphasised the creation of "positive energy" as part of an ideological campaign to encourage certain kinds of expression and limit others. Critics say the technology is based on a pseudo-science of stereotypes, and an increasing number of researchers, lawyers and rights activists believe it has serious implications for human rights, privacy and freedom of expression. With the global industry forecast to be worth nearly $36bn by 2023, growing at nearly 30% a year, rights groups say action needs to be taken now. The main office of Taigusys is tucked behind a few low-rise office buildings in Shenzhen. Visitors are greeted at the doorway by a series of cameras capturing their images on a big screen that displays body temperature, along with age estimates, and other statistics. Chen, a general manager at the company, says the system in the doorway is the company's bestseller at the moment because of high demand during the coronavirus pandemic. Chen hails emotion recognition as a way to predict dangerous behaviour by prisoners, detect potential criminals at police checkpoints, problem pupils in schools and elderly people experiencing dementia in care homes. Taigusys systems are installed in about 300 prisons, detention centres and remand facilities around China, connecting 60,000 cameras. "Violence and suicide are very common in detention centres," says Chen. "Even if police nowadays don't beat prisoners, they often try to wear them down by not allowing them to fall asleep.

The ACM Conference for Fairness, Accountability, and Transparency (FAccT) has decided to suspend its sponsorship relationship with Google, conference sponsorship co-chair and Boise State University assistant professor Michael Ekstrand confirmed today. The organizers of the AI ethics research conference came to this decision a little over a week after Google fired Ethical AI lead Margaret Mitchell and three months after the firing of Ethical AI co-lead Timnit Gebru. Google has subsequently reorganized about 100 engineers across 10 teams, including placing Ethical AI under the leadership of Google VP Marian Croak. "FAccT is guided by a Strategic Plan, and the conference by-laws charge the Sponsorship Chairs, in collaboration with the Executive Committee, with developing a sponsorship portfolio that aligns with that plan," Ekstrand told VentureBeat in an email. "The Executive Committee made the decision that having Google as a sponsor for the 2021 conference would not be in the best interests of the community and impede the Strategic Plan. We will be revising the sponsorship policy for next year's conference."

There have been predictions that Robots will take over our jobs. As of today, that prediction is rapidly coming to pass. Our imagination may tell us these robots are hardware, machines made of metal or carbon fibre. This is not quite the case, as these Robots are software called bots. Bots are programmed to repetitively automate operational and transactional tasks without the need for human input.

Machine learning has achieved tremendous success in a variety of domains in recent years. However, a lot of these success stories have been in places where the training and the testing distributions are extremely similar to each other. In everyday situations when models are tested in slightly different data than they were trained on, ML algorithms can fail spectacularly. This research attempts to formally define this problem, what sets of assumptions are reasonable to make in our data and what kind of guarantees we hope to obtain from them. Then, we focus on a certain class of out of distribution problems, their assumptions, and introduce simple algorithms that follow from these assumptions that are able to provide more reliable generalization. A central topic in the thesis is the strong link between discovering the causal structure of the data, finding features that are reliable (when using them to predict) regardless of their context, and out of distribution generalization.

Humans have a natural instinct to identify unknown object instances in their environments. The intrinsic curiosity about these unknown instances aids in learning about them, when the corresponding knowledge is eventually available. This motivates us to propose a novel computer vision problem called: `Open World Object Detection', where a model is tasked to: 1) identify objects that have not been introduced to it as `unknown', without explicit supervision to do so, and 2) incrementally learn these identified unknown categories without forgetting previously learned classes, when the corresponding labels are progressively received. We formulate the problem, introduce a strong evaluation protocol and provide a novel solution, which we call ORE: Open World Object Detector, based on contrastive clustering and energy based unknown identification. Our experimental evaluation and ablation studies analyze the efficacy of ORE in achieving Open World objectives. As an interesting by-product, we find that identifying and characterizing unknown instances helps to reduce confusion in an incremental object detection setting, where we achieve state-of-the-art performance, with no extra methodological effort. We hope that our work will attract further research into this newly identified, yet crucial research direction.

One crucial objective of multi-task learning is to align distributions across tasks so that the information between them can be transferred and shared. However, existing approaches only focused on matching the marginal feature distribution while ignoring the semantic information, which may hinder the learning performance. To address this issue, we propose to leverage the label information in multi-task learning by exploring the semantic conditional relations among tasks. We first theoretically analyze the generalization bound of multi-task learning based on the notion of Jensen-Shannon divergence, which provides new insights into the value of label information in multi-task learning. Our analysis also leads to a concrete algorithm that jointly matches the semantic distribution and controls label distribution divergence. To confirm the effectiveness of the proposed method, we first compare the algorithm with several baselines on some benchmarks and then test the algorithms under label space shift conditions. Empirical results demonstrate that the proposed method could outperform most baselines and achieve state-of-the-art performance, particularly showing the benefits under the label shift conditions.

We investigate upper bounds on the length of cost optimal plans that are valid for problems with 0-cost actions. We employ these upper bounds as horizons for a SAT-based encoding of planning with costs. Given an initial upper bound on the cost of the optimal plan, we experimentally show that this SAT-based approach is able to compute plans with better costs, and in many cases it can match the optimal cost. Also, in multiple instances, the approach is successful in proving that a certain cost is the optimal plan cost.

Throughout history, society has debated the morality of debt. In ancient times, debt--borrowing from another on the promise of repayment--was viewed in many cultures as sinful, with lending at interest especially repugnant. The concern that borrowers would become overindebted and enslaved to lenders meant that debts were routinely forgiven. These concerns continue to influence perceptions of lending and the regulation of credit markets today. Consider the prohibition against charging interest in Islamic finance and interest rate caps on payday lenders--companies that offer high-cost, short-term loans.

Konstantin Klemmer is a PhD student at the University of Warwick working at the intersection of machine learning and geographic data. He also serves as the Communications Chair for Climate Change AI. We talked about his research and the Climate Change AI organisation. Climate Change AI (CCAI) is a volunteer run organisation that catalyses impactful work at the intersection of climate change and machine learning by providing education and infrastructure, building a community, and advancing discourse. We also run a forum and regular community events like our fortnightly happy hour.