Verdict lists five of the top tweets on robotics in Q2 2022 based on data from GlobalData's Technology Influencer Platform. The top tweets are based on total engagements (likes and retweets) received on tweets from more than 375 robotics experts tracked by GlobalData's Technology Influencer platform during the second quarter (Q2) of 2022. Massimo, a technology expert, shared an article on the Chinese robot manufacturer Senpower building a self-transforming Transformer model, allowing them to convert between a standing toy and truck on its own. The makers of the Optimus Prime developed this concept and evolved it into the Robosen T9 robot toy, the article detailed. This version can walk, dance, drive, pose, and has 22 programmable servo motors that allows it to learn new skills.

Google experienced a major international internet outage on Tuesday, technology platforms reported. The realtime online platform Downdetector reported users had registered problems with Google explorer, the world's dominant search engine from 2.12am BST (9.12pm EST, 11.12AM AEST. As of 11.38AM, there had been 4,113 confirmed reports of Google outages. User reports indicate Google is having problems since 9:12 PM EDT. Users said sister platforms Gmail, Google maps and Google images were also experiencing problems.

Ripple's mission is to enable payments every way, everywhere for everyone. We believe connecting traditional financial entities like banks, payment providers and corporations with emerging blockchain technologies and users is the path to an open, decentralized, and more inclusive financial future. This Internet of Value gives any internet-enabled person, application or device access to financial services that are transparent, fast, reliable, and cheap. Delivering this vision is a challenge of massive scale spanning $155 trillion in annual cross border fiat payments and the $1.5 trillion market of digital assets that has grown 10X in the last year. We are looking for a Senior Software Engineer, Machine Learning Operations to join a new team charged with determining and delivering optimal liquidity for every customer in the world in a cost-effective, robust and scalable manner.

Error correction techniques remain effective to refine outputs from automatic speech recognition (ASR) models. Existing end-to-end error correction methods based on an encoder-decoder architecture process all tokens in the decoding phase, creating undesirable latency. In this paper, we propose an ASR error correction method utilizing the predictions of correction operations. More specifically, we construct a predictor between the encoder and the decoder to learn if a token should be kept ("K"), deleted ("D"), or changed ("C") to restrict decoding to only part of the input sequence embeddings (the "C" tokens) for fast inference. Experiments on three public datasets demonstrate the effectiveness of the proposed approach in reducing the latency of the decoding process in ASR correction. It enhances the inference speed by at least three times (3.4 and 5.7 times) while maintaining the same level of accuracy (with WER reductions of 0.53% and 1.69% respectively) for our two proposed models compared to a solid encoder-decoder baseline. In the meantime, we produce and release a benchmark dataset contributing to the ASR error correction community to foster research along this line.

Conducting randomized experiments in education settings raises the question of how we can use machine learning techniques to improve educational interventions. Using Multi-Armed Bandits (MAB) algorithms like Thompson Sampling (TS) in adaptive experiments can increase students' chances of obtaining better outcomes by increasing the probability of assignment to the most optimal condition (arm), even before an intervention completes. This is an advantage over traditional A/B testing, which may allocate an equal number of students to both optimal and non-optimal conditions. The problem is the exploration-exploitation trade-off. Even though adaptive policies aim to collect enough information to allocate more students to better arms reliably, past work shows that this may not be enough exploration to draw reliable conclusions about whether arms differ. Hence, it is of interest to provide additional uniform random (UR) exploration throughout the experiment. This paper shows a real-world adaptive experiment on how students engage with instructors' weekly email reminders to build their time management habits. Our metric of interest is open email rates which tracks the arms represented by different subject lines. These are delivered following different allocation algorithms: UR, TS, and what we identified as TS{\dag} - which combines both TS and UR rewards to update its priors. We highlight problems with these adaptive algorithms - such as possible exploitation of an arm when there is no significant difference - and address their causes and consequences. Future directions includes studying situations where the early choice of the optimal arm is not ideal and how adaptive algorithms can address them.

Like the European Union (EU)'s General Data Protection Regulation (GDPR) that entered into force in 2016, the upcoming Artificial Intelligence (AI) Act will have extraterritorial scope and global impact. Considering the AI Act's broad scope and the financial risks relating to non-compliance, businesses must prepare for these future regulatory changes now and proactively take the initiatives to comply with best practices early on, according to a new whitepaper by Swiss data services company Unit8. The paper, titled Upcoming AI Regulation: What to expect and how to prepare, delves into the EU's forthcoming AI Act, providing insights into the future development of AI regulation in Europe and the potential implications for organizations worldwide. The European Commission (EC) unveiled a proposal for a legal framework on AI in April 2021, seeking to address risks of specifically created by AI applications, proposing a list of high risk applications, setting clear requirements for AI systems for high risk applications and defining specific obligations for AI users and providers of high risk applications. The proposed rules also propose a conformity assessment method for AI systems, propose enforcement after an AI system is placed in the market, and propose a governance structure at European and national level.

Drones equipped with cameras can significantly enhance human ability to perceive the world because of their remarkable maneuverability in 3D space. Ironically, object detection for drones has always been conducted in the 2D image space, which fundamentally limits their ability to understand 3D scenes. Furthermore, existing 3D object detection methods developed for autonomous driving cannot be directly applied to drones due to the lack of deformation modeling, which is essential for the distant aerial perspective with sensitive distortion and small objects. To fill the gap, this work proposes a dual-view detection system named DVDET to achieve aerial monocular object detection in both the 2D image space and the 3D physical space. To address the severe view deformation issue, we propose a novel trainable geo-deformable transformation module that can properly warp information from the drone's perspective to the BEV. Compared to the monocular methods for cars, our transformation includes a learnable deformable network for explicitly revising the severe deviation. To address the dataset challenge, we propose a new large-scale simulation dataset named AM3D-Sim, generated by the co-simulation of AirSIM and CARLA, and a new real-world aerial dataset named AM3D-Real, collected by DJI Matrice 300 RTK, in both datasets, high-quality annotations for 3D object detection are provided. Extensive experiments show that i) aerial monocular 3D object detection is feasible; ii) the model pre-trained on the simulation dataset benefits real-world performance, and iii) DVDET also benefits monocular 3D object detection for cars. To encourage more researchers to investigate this area, we will release the dataset and related code in https://sjtu-magic.github.io/dataset/AM3D/.

While large pretrained Transformer models have proven highly capable at tackling natural language tasks, handling long sequence inputs continues to be a significant challenge. One such task is long input summarization, where inputs are longer than the maximum input context of most pretrained models. Through an extensive set of experiments, we investigate what model architectural changes and pretraining paradigms can most efficiently adapt a pretrained Transformer for long input summarization. We find that a staggered, block-local Transformer with global encoder tokens strikes a good balance of performance and efficiency, and that an additional pretraining phase on long sequences meaningfully improves downstream summarization performance. Based on our findings, we introduce PEGASUS-X, an extension of the PEGASUS model with additional long input pretraining to handle inputs of up to 16K tokens. PEGASUS-X achieves strong performance on long input summarization tasks comparable with much larger models while adding few additional parameters and not requiring model parallelism to train.

Non-parallel many-to-many voice conversion remains an interesting but challenging speech processing task. Recently, AutoVC, a conditional autoencoder based method, achieved excellent conversion results by disentangling the speaker identity and the speech content using information-constraining bottlenecks. However, due to the pure autoencoder training method, it is difficult to evaluate the separation effect of content and speaker identity. In this paper, a novel voice conversion framework, named $\boldsymbol T$ext $\boldsymbol G$uided $\boldsymbol A$utoVC(TGAVC), is proposed to more effectively separate content and timbre from speech, where an expected content embedding produced based on the text transcriptions is designed to guide the extraction of voice content. In addition, the adversarial training is applied to eliminate the speaker identity information in the estimated content embedding extracted from speech. Under the guidance of the expected content embedding and the adversarial training, the content encoder is trained to extract speaker-independent content embedding from speech. Experiments on AIShell-3 dataset show that the proposed model outperforms AutoVC in terms of naturalness and similarity of converted speech.

Deep and wide neural networks successfully fit very complex functions today, but dense models are starting to be prohibitively expensive for inference. To mitigate this, one promising direction is networks that activate a sparse subgraph of the network. The subgraph is chosen by a data-dependent routing function, enforcing a fixed mapping of inputs to subnetworks (e.g., the Mixture of Experts (MoE) paradigm in Switch Transformers). However, prior work is largely empirical, and while existing routing functions work well in practice, they do not lead to theoretical guarantees on approximation ability. We aim to provide a theoretical explanation for the power of sparse networks. As our first contribution, we present a formal model of data-dependent sparse networks that captures salient aspects of popular architectures. We then introduce a routing function based on locality sensitive hashing (LSH) that enables us to reason about how well sparse networks approximate target functions. After representing LSH-based sparse networks with our model, we prove that sparse networks can match the approximation power of dense networks on Lipschitz functions. Applying LSH on the input vectors means that the experts interpolate the target function in different subregions of the input space. To support our theory, we define various datasets based on Lipschitz target functions, and we show that sparse networks give a favorable trade-off between number of active units and approximation quality.