Overnight Russian drone attacks on east, south and central Ukraine damaged civilian infrastructure and businesses in the Poltava region and injured civilians in the Odesa region, Ukrainian officials said early on Wednesday. Odesa came under a "massive attack" by Russian drones overnight on Tuesday, wounding at least three people, the head of the regional administration, Oleh Kiper, wrote on his Telegram page. A residential building in a densely populated urban area of Odesa, civilian infrastructure and an educational facility were hit, he said. Air defence units repelled Russian air attacks on the Kyiv region and Ukraine's second largest city of Kharkiv, regional governors said in posts on Telegram channels. Russian forces said they have retaken St Nicholas Belogorsky monastery in the village of Gornal in Russia's Kursk region, where Ukrainian troops had been based, Russia's TASS news agency quoted a security source as saying.

At least 51 people were killed and 271 injured when two Russian ballistic missiles hit a military academy and a nearby hospital in Ukraine's central town of Poltava in the deadliest single attack in the war this year. The missiles hit shortly after the air raid alert sounded when many people were on their way to a bomb shelter, the Ministry of Defence said in a statement. A woman and her eight-year-old son were killed after the Zaporizhia region hotel where they were staying was hit in a Russian missile attack. Her husband and daughter were injured, Ukrainian authorities said. Ukraine's Ministry of Energy said three employees were injured in a Russian drone attack on a power facility in the northern region of Chernihiv.

When a speaker verification (SV) system operates far from the sound sourced, significant challenges arise due to the interference of noise and reverberation. Studies have shown that incorporating phonetic information into speaker embedding can improve the performance of text-independent SV. Inspired by this observation, we propose a joint-training speech recognition and speaker recognition (JTSS) framework to exploit phonetic content for far-field SV. The framework encourages speaker embeddings to preserve phonetic information by matching the frame-based feature maps of a speaker embedding network with wav2vec's vectors. The intuition is that phonetic information can preserve low-level acoustic dynamics with speaker information and thus partly compensate for the degradation due to noise and reverberation. Results show that the proposed framework outperforms the standard speaker embedding on the VOiCES Challenge 2019 evaluation set and the VoxCeleb1 test set. This indicates that leveraging phonetic information under far-field conditions is effective for learning robust speaker representations.

Stochastic collocation (SC) is a well-known non-intrusive method of constructing surrogate models for uncertainty quantification. In dynamical systems, SC is especially suited for full-field uncertainty propagation that characterizes the distributions of the high-dimensional primary solution fields of a model with stochastic input parameters. However, due to the highly nonlinear nature of the parameter-to-solution map in even the simplest dynamical systems, the constructed SC surrogates are often inaccurate. This work presents an alternative approach, where we apply the SC approximation over the dynamics of the model, rather than the solution. By combining the data-driven sparse identification of nonlinear dynamics (SINDy) framework with SC, we construct dynamics surrogates and integrate them through time to construct the surrogate solutions. We demonstrate that the SC-over-dynamics framework leads to smaller errors, both in terms of the approximated system trajectories as well as the model state distributions, when compared against full-field SC applied to the solutions directly.

Ukraine and Russia launched waves of drone attacks overnight with reports of a fire at an oil refinery in Ukraine's Poltava region and four Ukrainian unmanned aerial vehicles (UAVs) being shot down over two regions in Russia's west, officials say. A Russian drone hit the Kremenchuk oil refinery in the central Poltava region of Ukraine, causing a fire, the regional governor, Dmytro Lunin, said on Wednesday. "Last night, Russians repeatedly attacked Poltava region. Our air defence system did a good job against enemy UAVs," he said on the Telegram messaging app. The General Staff of Ukraine's Armed Forces said air defence systems shot down 17 of 24 drones that Russia launched against targets in Ukraine.

Russia has launched air attacks on targets in southern and eastern Ukraine using drones and possibly ballistic missiles, Ukraine's air force said. The southern port of Odesa and the Mykolaiv, Donetsk, Kherson, Zaporizhia and Dnipropetrovsk regions were all under threat of Russian drone attacks, the air force said on the Telegram messaging app in the early hours of Tuesday morning. Russia may also be using ballistic weaponry to attack the regions of Poltava, Cherkasy, Dnipropetrovsk, Kharkiv and Kirovohrad, the air force added. Oleh Kiper, the head of the Odesa region's military administration, said air defence systems there were engaged in repelling a Russian drone attack. "Several waves of attacks are likely," Kiper said on Telegram.

In this paper, we study the problem of inferring spatially-varying Gaussian Markov random fields (SV-GMRF) where the goal is to learn a network of sparse, context-specific GMRFs representing network relationships between genes. An important application of SV-GMRFs is in inference of gene regulatory networks from spatially-resolved transcriptomics datasets. The current work on inference of SV-GMRFs are based on the regularized maximum likelihood estimation (MLE) and suffer from overwhelmingly high computational cost due to their highly nonlinear nature. To alleviate this challenge, we propose a simple and efficient optimization problem in lieu of MLE that comes equipped with strong statistical and computational guarantees. Our proposed optimization problem is extremely efficient in practice: we can solve instances of SV-GMRFs with more than 2 million variables in less than 2 minutes. We apply the developed framework to study how gene regulatory networks in Glioblastoma are spatially rewired within tissue, and identify prominent activity of the transcription factor HES4 and ribosomal proteins as characterizing the gene expression network in the tumor peri-vascular niche that is known to harbor treatment resistant stem cells.