Heat-assisted detection and ranging.

Machine perception uses advanced sensors to collect information about the surrounding scene for situational awareness1-7. State-of-the-art machine perception8 using active sonar, radar and LiDAR to enhance camera vision9 faces difficulties when the number of intelligent agents scales up10,11. Exploiting omnipresent heat signal could be a new frontier for scalable perception. However, objects and their environment constantly emit and scatter thermal radiation, leading to textureless images famously known as the 'ghosting effect'12. Thermal vision thus has no specificity limited by information loss, whereas thermal ranging-crucial for navigation-has been elusive even when combined with artificial intelligence (AI)13. Here we propose and experimentally demonstrate heat-assisted detection and ranging (HADAR) overcoming this open challenge of ghosting and benchmark it against AI-enhanced thermal sensing. HADAR not only sees texture and depth through the darkness as if it were day but also perceives decluttered physical attributes beyond RGB or thermal vision, paving the way to fully passive and physics-aware machine perception. We develop HADAR estimation theory and address its photonic shot-noise limits depicting information-theoretic bounds to HADAR-based AI performance. HADAR ranging at night beats thermal ranging and shows an accuracy comparable with RGB stereovision in daylight. Our automated HADAR thermography reaches the Cramér-Rao bound on temperature accuracy, beating existing thermography techniques. Our work leads to a disruptive technology that can accelerate the Fourth Industrial Revolution (Industry 4.0)14 with HADAR-based autonomous navigation and human-robot social interactions.

Quantum-accelerated imaging of N stars.

Imaging point sources with low angular separation near or below the Rayleigh criterion are important in astronomy, e.g., in the search for habitable exoplanets near stars. However, the measurement time required to resolve stars in the sub-Rayleigh region via traditional direct imaging is usually prohibitive. Here we propose quantum-accelerated imaging (QAI) to significantly reduce the measurement time using an information-theoretic approach. QAI achieves quantum acceleration by adaptively learning optimal measurements from data to maximize Fisher information per detected photon. Our approach can be implemented experimentally by linear-projection instruments followed by single-photon detectors. We estimate the position, brightness, and the number of unknown stars 10∼100 times faster than direct imaging with the same aperture. QAI is scalable to a large number of incoherent point sources and can find widespread applicability beyond astronomy to high-speed imaging, fluorescence microscopy, and efficient optical read-out of qubits.

Machine Learning for Communications.

Due to the proliferation of applications and services that run over communication networks, ranging from video streaming and data analytics to robotics and augmented reality, tomorrow's networks will be faced with increasing challenges resulting from the explosive growth of data traffic demand with significantly varying performance requirements [...].

Scheduling and Power Control for Wireless Multicast Systems via Deep Reinforcement Learning.

Multicasting in wireless systems is a natural way to exploit the redundancy in user requests in a content centric network. Power control and optimal scheduling can significantly improve the wireless multicast network's performance under fading. However, the model-based approaches for power control and scheduling studied earlier are not scalable to large state spaces or changing system dynamics. In this paper, we use deep reinforcement learning, where we use function approximation of the Q-function via a deep neural network to obtain a power control policy that matches the optimal policy for a small network. We show that power control policy can be learned for reasonably large systems via this approach. Further, we use multi-timescale stochastic optimization to maintain the average power constraint. We demonstrate that a slight modification of the learning algorithm allows tracking of time varying system statistics. Finally, we extend the multi-time scale approach to simultaneously learn the optimal queuing strategy along with power control. We demonstrate the scalability, tracking and cross-layer optimization capabilities of our algorithms via simulations. The proposed multi-time scale approach can be used in general large state-space dynamical systems with multiple objectives and constraints, and may be of independent interest.

A computer vision approach for classifying isometric grip force exertion levels.

Exposure to high and/or repetitive force exertions can lead to musculoskeletal injuries. However, measuring worker force exertion levels is challenging, and existing techniques can be intrusive, interfere with human-machine interface, and/or limited by subjectivity. In this work, computer vision techniques are developed to detect isometric grip exertions using facial videos and wearable photoplethysmogram. Eighteen participants (19-24 years) performed isometric grip exertions at varying levels of maximum voluntary contraction. Novel features that predict forces were identified and extracted from video and photoplethysmogram data. Two experiments with two (High/Low) and three (0%MVC/50%MVC/100%MVC) labels were performed to classify exertions. The Deep Neural Network classifier performed the best with 96% and 87% accuracy for two- and three-level classifications, respectively. This approach was robust to leave subjects out during cross-validation (86% accuracy when 3-subjects were left out) and robust to noise (i.e. 89% accuracy for correctly classifying talking activities as low force exertions). Practitioner summary: Forceful exertions are contributing factors to musculoskeletal injuries, yet it remains difficult to measure in work environments. This paper presents an approach to estimate force exertion levels, which is less distracting to workers, easier to implement by practitioners, and could potentially be used in a wide variety of workplaces. Abbreviations: MSD: musculoskeletal disorders; ACGIH: American Conference of Governmental Industrial Hygienists; HAL: hand activity level; MVC: maximum voluntary contraction; PPG: photoplethysmogram; DNN: deep neural networks; LOSO: leave-one-subject-out; ROC: receiver operating characteristic; AUC: area under curve.

Improving Molecule Generation and Drug Discovery with a Knowledge-enhanced Generative Model.

Recent advancements in generative models have established state-of-the-art benchmarks in the generation of molecules and novel drug candidates. Despite these successes, a significant gap persists between generative models and the utilization of extensive biomedical knowledge, often systematized within knowledge graphs, whose potential to inform and enhance generative processes has not been realized. In this paper, we present a novel approach that bridges this divide by developing a framework for knowledge-enhanced generative models called K-DReAM. We develop a scalable methodology to extend the functionality of knowledge graphs while preserving semantic integrity, and incorporate this contextual information into a generative framework to guide a diffusion-based model. The integration of knowledge graph embeddings with our generative model furnishes a robust mechanism for producing novel drug candidates possessing specific characteristics while ensuring validity and synthesizability. K-DReAM outperforms state-of-the-art generative models on both unconditional and targeted generation tasks.

Understanding the Natural Language of DNA using Encoder-Decoder Foundation Models with Byte-level Precision.

This paper presents the Ensemble Nucleotide Byte-level Encoder-Decoder (ENBED) foundation model, analyzing DNA sequences at byte-level precision with an encoder-decoder Transformer architecture. ENBED uses a sub-quadratic implementation of attention to develop an efficient model capable of sequence-to-sequence transformations, generalizing previous genomic models with encoder-only or decoder-only architectures. We use Masked Language Modeling to pre-train the foundation model using reference genome sequences and apply it in the following downstream tasks: (1) identification of enhancers, promotors and splice sites, (2) recognition of sequences containing base call mismatches and insertion/deletion errors, an advantage over tokenization schemes involving multiple base pairs, which lose the ability to analyze with byte-level precision, (3) identification of biological function annotations of genomic sequences, and (4) generating mutations of the Influenza virus using the encoder-decoder architecture and validating them against real-world observations. In each of these tasks, we demonstrate significant improvement as compared to the existing state-of-the-art results.

Diazinon-chemistry and environmental fate: a California perspective.

Diazinon, first introduced in USA in 1956, is a broad-spectrum contact organophosphate pesticide that has been used as an insecticide, and nematicide. It has been ond of the most widely used insecticides in the USA for household and agricultural pest control. In 2004, residential use of diazinon was discontinued; as a result, the total amount applied has drastically decreased. [corrected]. Consequently, the amounts of diazinon applied have been drastically decreased. For example, in California, the amount of diazinon applied decreased from 501,784 kg in 2000 to 64,122 kg in 2010. Diazinon has a K(oc) value of 40-432 and is considered to be moderately mobile in soils. Diazinon residues have been detected in groundwater, drinking water wells, monitoring wells, and agricultural well. The highest detection frequencies and highest percentages of exceedance of the water quality criterion value of 0.1 µg/L have been reported from the top five agricultural counties n California that had the highest diazinon use. Diazinon is transported in air via atmospheric processes such as direct air movement and wet deposition in snow and rain, although concentrations decrease with distance and evaluation from the source. In the environment, diazinon undergoes degradation by several processes, the most important of which is microbial degradation in soils. The rate of diazinon degradation is affected by pH, soil type, organic amendments, soil moisture, and the concentration of diazinon in the soil, with soil pH being a major influencing factor in diazinon degradation rate. Studies indicate tha soil organic matter is the most important factor that influences diazinon sorption by soils, although clay content and soil ph also play an important role in diazinon sorption. Diazinon is very highly to moderately toxic aquatic arganisms, Diazinon inhibits the enzyme acetylcholinesterase, which hydrolyzes the neurotransmitter acetylcholine and leads to a suite of intermediate syndromes including anorexia, diarrhea, generalized weakness, muscle tremors, abnormal posturing and behavior, depression, and health. Differences in metabolism among species and exposure concentrations play a vital role in diazinon's bioaccumulation among different aquatic organisms in a wide range of accumulating rates and efficiencies.

Hierarchical Adversarial Inverse Reinforcement Learning.

Imitation learning (IL) has been proposed to recover the expert policy from demonstrations. However, it would be difficult to learn a single monolithic policy for highly complex long-horizon tasks of which the expert policy usually contains subtask hierarchies. Therefore, hierarchical IL (HIL) has been developed to learn a hierarchical policy from expert demonstrations through explicitly modeling the activity structure in a task with the option framework. Existing HIL methods either overlook the causal relationship between the subtask structure and the learned policy, or fail to learn the high-level and low-level policy in the hierarchical framework in conjuncture, which leads to suboptimality. In this work, we propose a novel HIL algorithm-hierarchical adversarial inverse reinforcement learning (H-AIRL), which extends a state-of-the-art (SOTA) IL algorithm-AIRL, with the one-step option framework. Specifically, we redefine the AIRL objectives on the extended state and action spaces, and further introduce a directed information term to the objective function to enhance the causality between the low-level policy and its corresponding subtask. Moreover, we propose an expectation-maximization (EM) adaption of our algorithm so that it can be applied to expert demonstrations without the subtask annotations which are more accessible in practice. Theoretical justifications of our algorithm design and evaluations on challenging robotic control tasks are provided to show the superiority of our algorithm compared with SOTA HIL baselines. The codes are available at https://github.com/LucasCJYSDL/HierAIRL.

Tensor learning of pointwise mutual information from EHR data for early prediction of sepsis.

Early detection of sepsis can facilitate early clinical intervention with effective treatment and may reduce sepsis mortality rates. In view of this, machine learning-based automated diagnosis of sepsis using easily recordable physiological data can be more promising as compared to the gold standard rule-based clinical criteria in current practice. This study aims to develop such a machine learning framework that demonstrates the quantification of heterogeneity within the tabular electronic health records (EHR) data of clinical covariates to capture both linear relationships and nonlinear correlation for the early prediction of sepsis. Here, the statistics of pairwise association for each hour-covariate pair within the EHR data for every 6-hours window-duration with selected 24 covariates is described using pointwise mutual information (PMI) matrix. This matrix gives the heterogeneity of data as a two-dimensional map. Such matrices are fused horizontally along the z-axis as vertical slices in the xy plane to form a 3-way tensor for each record with the corresponding Length of Stay (L). Tensor factorization of such fused tensor for every record is performed using Tucker decomposition, and only the core tensors are retained later, excluding the 3 unitary matrices to provide the latent feature set for the prediction of sepsis onset. A five-fold cross-validation scheme is employed wherein the obtained 120 latent features from the reshaped core tensor, are fed to Light Gradient Boosting Machine Learning models (LightGBM) for binary classification, further alleviating the involved class imbalance. The machine-learning framework is designed via Bayesian optimization, yielding an average normalized utility score of 0.4519 as defined by challenge organizers and area under the receiver operating characteristic curve (AUROC) of 0.8621 on publicly available PhysioNet/Computing in Cardiology Challenge 2019 training data. The proposed tensor decomposition of 3-way fused tensor formulated using PMI matrices leverages higher-order temporal interactions between the pairwise associations among the clinical values for early prediction of sepsis. This is validated with improved risk prediction power for every hour of admission to the ICU in terms of utility score, AUROC, and F1 score. The results obtained show a significant improvement particularly in terms of utility score of ~1.5-2% under a 5-fold cross-validation scheme on entire training data as compared to a top entrant research study that participated in the challenge.

Triazine adsorption by saponite and beidellite clay minerals.

We investigated the sorption of three triazine herbicides (atrazine, simazine, and metribuzin) by saponite and beidellite clay minerals saturated with K+, Cs+, Na+, and Ca2+. Saponite clay sorbed a larger fraction of each pesticide from aqueous solution than did beidellite clay. The lower cation-exchange capacity in saponite compared to that in beidellite presumably results in a less crowded interlayer, with more siloxane surface being available for adsorption. Generally, Cs-saturated clays sorbed more triazines than did clays saturated by K+, Na+, or Ca2+. We attribute this to the smaller hydrated radius of Cs+, which again increases the siloxane surface that is available for sorption. Furthermore, the relatively weak hydration of Cs+ reduces the swelling of clay interlayers, thus making sorption domains less hydrated and more receptive to hydrophobic molecules. The Cs-saponite manifested a sorption of more than 1% atrazine by weight above equilibrium concentrations of 6 mg/L, which to our knowledge is the largest sorption of neutral atrazine from water yet reported for an inorganic sorbent. Molecular dynamics simulations indicate that atrazine interacts both with clay basal planes and with multiple cations in the clay interlayer.

Enhanced sorption of trichloroethene by smectite clay exchanged with Cs+.

Trichloroethene (TCE) is one of the most common pollutants in groundwater, and Cs+ can be a cocontaminant at nuclear facilities. Smectite clays have large surface areas, are common in soils, have high affinities for some organic contaminants, and hence can potentially influence the transport of organic pollutants entering soils and sediments. The exchangeable cations present near smectite clay surfaces can radically influence the sorption of organic pollutants by soil clays. This research was undertaken to determine the effect of Cs+, and other common interlayer cations, such as K+ and Ca2+, on the sorption of TCE by a reference smectite clay saponite. Cs-saturated clay sorbed the most TCE, up to 3500 mg/kg, while Ca-saturated smectite sorbed the least. We hypothesize that the stronger sorption of TCE by the Cs-smectite can be attributed to the lower hydration energy and hence smaller hydrated radius of Cs+, which expands the lateral clay surface domains available for sorption. Also, Cs-smectite interlayers are only one or two water layers thick, which may drive capillary condensation of TCE. Our results implicate enhanced retention of TCE in aquifer materials containing smectites accompanied by Cs+ cocontamination.

Influence of smectite hydration and swelling on atrazine sorption behavior.

Smectites, clay minerals commonly found in soils and sediments, vary widely in their ability to adsorb organic chemicals. Recent research has demonstrated the importance of surface charge density and properties of exchangeable cations in controlling the affinity of smectites for organic molecules. In this study, we induced hysteresis in the crystalline swelling of smectites to test the hypothesis that the extent of crystalline swelling (or interlayer hydration status) has a large influence on the ability of smectites to adsorb atrazine from aqueous systems. Air-dried K-saturated Panther Creek (PC) smectite swelled less (d(001) = 1.38 nm) than never-dried K-PC (d(001) = 1.7 nm) when rehydrated in 20 mM KCl. Correspondingly, the air-dried-rehydrated K-PC had an order of magnitude greater affinity for atrazine relative to the never-dried K-PC. Both air-dried-rehydrated and never-dried Ca-PC expanded to approximately 2.0 nm in 10 mM CaCl2 and both samples had similar affinities for atrazine that were slightly lower than that of never-dried K-PC. The importance of interlayer hydration status in controlling sorption affinity was confirmed by molecular modeling, which revealed much greater interaction between interlayer water molecules and atrazine in a three-layer hydrate relative to a one-layer hydrate. The entropy change on moving atrazine from a fully hydrated state in the bulk solution to a partially hydrated state in the smectite interlayers is believed to be a major factor influencing sorption affinity. In an application test, choice of background solution (20 mM KCl versus 10 mM CaCl2) and air-drying treatments significantly affected atrazine sorption affinities for three-smectitic soils; however, the trends were not consistent with those observed for the reference smectite. Further, extending the initial rehydration time from 24 to 240 h (prior to adding atrazine) significantly decreased the soil's sorption affinity for atrazine. We conclude that interlayer hydration status has a large influence on the affinity of smectites for atrazine and that air-drying treatments have the potential to modify the sorption affinity of smectitic soils for organic molecules such as atrazine.