Retrieval Framework Research Articles

Clouds form on the hot Saturn JWST ERO target WASP-96b

Context. WASP-96b is a hot Saturn exoplanet, with an equilibrium temperature of ≈1300 K. This is well within the regime of thermo-dynamically expected extensive cloud formation. Prior observations with Hubble/WFC3, Spitzer/IRAC, and VLT/FORS2 have been combined into a single spectrum for which retrievals suggest a cold but cloud-free atmosphere. Recently, the planet was observed with the James Webb Space Telescope (JWST) as part of the Early Release Observations (ERO). Aims. The formation of clouds in the atmosphere of exoplanet WASP-96b is explored. Methods. One-dimensional profiles were extracted from the 3D GCM expeRT/MITgcm results and used as input for a kinetic, non-equilibrium model to study the formation of mineral cloud particles of mixed composition. The ARCiS retrieval framework was applied to the pre-JWST WASP-96b transit spectrum to investigate the apparent contradiction between cloudy models and assumed cloud-free transit spectrum. Results. Clouds are predicted to be ubiquitous throughout the atmosphere of WASP-96b. Silicate materials contribute between 40% and 90% cloud particle volume, which means that metal oxides also contribute with up to 40% cloud particle volume in the low-pressure regimes that affect spectra. We explore how these cloudy models match currently available transit spectra. Reduced vertical mixing acts to settle clouds to deeper in the atmosphere, and an increased cloud particle porosity reduces the opacity of clouds in the near-IR and optical region. These two effects allow for clearer molecular features to be observed while still allowing clouds to be in the atmosphere. Conclusions. The atmosphere of WASP-96b is unlikely to be cloud free. Retrievals of HST, Spitzer, and VLT spectra also show that multiple cloudy solutions reproduce the data. JWST observations will be affected by clouds, where the cloud top pressure varies by an order of magnitude within even the NIRISS wavelength range. The long-wavelength end of NIRSpec and the short-wavelength end of MIRI may probe atmospheric asymmetries between the limbs of the terminator on WASP-96b.

Proximity-Aware Clinical Passage Retrieval Framework by Exploiting Knowledge Structure

Clinicians have minimal time to search for and absorb the information needed while performing the duties of their medical practice. Their time-pressured situations requires the relevant information in search to be retrieved and presented in a more succinct form, such as in a short passage, rather than a whole page or document. In this context, clinical decision support (CDS) searches are beneficial when used to retrieve critical medical passages that can assist the practice of medical experts by offering appropriate medical information relevant to the medical case at hand. We present a novel CDS search framework designed for passage retrieval in order to support clinical decision-making using laboratory test results by incorporating proximity information. To do so, we use a knowledge structure that graphically visualizes key concepts and the corresponding relationships in a specific domain where nodes denote with their associative relationships. Furthermore, unlike previous studies that exploit knowledge structures during a re-ranking step, only dealing with initially highly retrieved passages, we utilize a knowledge structure for the purpose of query expansion. By doing so, our approach can unveil passages that are not retrieved during the initial retrieval process by including latent terms in the query list. We compared two models with/without edge pruning to capture a more latent relationship between terms. Our experiment showed that the embedded-based knowledge structure outperformed previous knowledge structure building approaches and other proximity-aware state-of-the-art models.

Context Parameters Retrieval Framework from Electronic Healthcare Record through Biomedical NLP for Clinical Support

Context Parameters Retrieval Framework from Electronic Healthcare Record through Biomedical NLP for Clinical Support

A Cloud-Dependent 1DVAR Precipitation Retrieval Algorithm for FengYun-3D Microwave Soundings: A Case Study in Tropical Cyclone Mekkhala

A cloud-dependent 1-D variational (1DVAR) precipitation retrieval algorithm applied to FengYun-3D microwave soundings (CD1DVAR-FY3DMS) is developed in this study. Compared with the current 1DVAR precipitation retrieval framework, cloud scene identification (CSI) is first proposed to delineate the different weather conditions including clear sky, stratiform clouds, and convective clouds. Results of this study in a typical tropical cyclone event demonstrate that: 1) the precipitation retrievals considering CSI (correlation coefficient ~0.72 and root mean square error ~1.63 mm/h) outperform those without distinguishing cloud scenes (correlation coefficient ~0.66 and root mean square error ~1.82 mm/h); 2) identifying cloud scenes in a variational scheme could significantly improve the accuracy of retrieving heavy precipitation volumes, with the capturing abilities improved by ~100%; and 3) the FengYun-3 constellation has great potential to complement global microwave retrievals. In addition, these findings could provide valuable references and pathfinders for further improving the retrieval accuracy of microwave-based precipitation estimates, especially for strong convective zones.

Context parameters retrieval framework from electronic healthcare record through biomedical NLP for clinical support

This paper presents architecture for extracting context features from electronic health records (EHR) to design a clinical support system using transfer learning for natural language processing (NLP). The system is trained to provide a supporting summary to the medical practioners on the basis of the ICD 9 codes and respective symptoms of the patient. The BioALBERT model has been trained over biomedical corpora and the proposed model uses improvised parameter sharing techniques and requires less physical memory. The theoretical analysis of the proposed system is supported by the experimental analysis. MIMIC-III database has been used to fine train the proposed models and to assess the efficiency and efficacy of the proposed work. This study introduced a context-aware approach for extracting useful context from EHR, which can be used to acquire a basic understanding of the treatment path.

DAKRS: Domain Adaptive Knowledge-Based Retrieval System for Natural Language-Based Vehicle Retrieval

Given Natural Language (NL) text descriptions, NL-based vehicle retrieval aims to extract target vehicles from a multi-view multi-camera traffic video pool. Due to inherent distinctions between textual and visual data, this is a challenging multi-modal retrieval task that requires robust feature extractors (e.g. neural network) to well-align the abstract representations of texts and images in the same domain. However, solutions to the problem have been challenged by the high data complexities of not only the multi-view, multi-camera attributes of visual data and the diverse range of textual descriptions but also a lack of high-volume datasets in this relatively new field, alongside a prominently large domain gap between training and test sets. Many existing approaches have developed computationally expensive models to separately extract the subspaces of language and vision before blending into the same shared representation space while only focusing on single-modal information and ignoring much of the multi-modal information to deal with the aforementioned issues. Hence, we propose a Domain Adaptive Knowledge-based Retrieval System (DAKRS) to effectively and efficiently align multi-modal knowledge in a setting of limited labels. Our contributions are threefold: (i) An efficient extension of Contrastive Language-Image Pre-training (CLIP)’s transfer learning into a baseline text-to-image multi-modular vehicle retrieval framework; (ii) A data enhancement module to create pseudo-vehicle tracks from the traffic video pool by leveraging the robustness of baseline retrieval model combine with background subtraction; and (iii) A SSDA (SSDA) scheme to engineer pseudo-labels for adapting model parameters to the target domain distribution. Experimental results are benchmarked on the Cityflow-NL dataset, illustrating our competitiveness against state-of-the-art performances in terms of effectiveness and efficiency without needing further post-processing or ensembling.

Cross-camera Trajectories Help Person Retrieval in a Camera Network.

We are concerned with retrieving a query person from multiple videos captured by a non-overlapping camera network. Existing methods often rely on purely visual matching or consider temporal constraints but ignore the spatial information of the camera network. To address this issue, we propose a pedestrian retrieval framework based on cross-camera trajectory generation that integrates both temporal and spatial information. To obtain pedestrian trajectories, we propose a novel cross-camera spatio-temporal model that integrates pedestrians' walking habits and the path layout between cameras to form a joint probability distribution. Such a cross-camera spatio-temporal model can be specified using sparsely sampled pedestrian data. Based on the spatio-temporal model, cross-camera trajectories can be extracted by the conditional random field model and further optimised by restricted non-negative matrix factorization. Finally, a trajectory re-ranking technique is proposed to improve the pedestrian retrieval results. To verify the effectiveness of our method, we construct the first cross-camera pedestrian trajectory dataset, the Person Trajectory Dataset, in real surveillance scenarios. Extensive experiments verify the effectiveness and robustness of the proposed method.

OMCBIR: Offline mobile content-based image retrieval with lightweight CNN optimization

Convolutional Neural Networks (CNNs) have achieved great success in computer vision applications. However, due to the high requirements for computation power and memory usage, most state-of-the-art CNNs are difficult to deploy on resource-constrained mobile devices. Although many typical lightweight neural networks have been proposed in the industry, such as MobileNetV2, which reduce the amount of parameters and calculations, they still have a lot of redundancy. Furthermore, few papers consider the use of deep learning models to implement image retrieval on terminals, so we propose a new offline retrieval framework based on lightweight neural network models, called Offline Mobile Content-Based Image Retrieval (OMCBIR). In this framework, we focus on the feature extraction of the model, by introducing pointwise group convolution and channel shuffle into the bottleneck block, reconstructing the network structure, and introducing the convolutional attention module, we propose an extremely lightweight small network-Attention-based Lightweight Network (ALNet). Compared to MobileNetV2, ALNet obtains a higher mAP on each dataset in OMCBIR when the model parameters are reduced by more than 62% and the model size is reduced by more than 63%. Extensive experiments conducted on five public datasets provide a trade-off between retrieval performance and model size of different algorithms, which proves the efficiency of the proposed OMCBIR.

A High-Dimensional Sparse Hashing Framework for Cross-Modal Retrieval

In recent years, many achievements have been made in improving the performance of supervised cross-modal hashing. However, it remains an open issue on how to fully explore the data information to achieve fine-grained retrieval performance. Most methods employ logical labels or a binary similarity matrix to supervise the hash learning, losing a lot of useful information. From another point of view, the low expressiveness of dense hash code severely limits its preservation of fine-grained data information. With this motivation, in this paper, we propose a high-dimensional sparse hashing framework for cross-modal retrieval, i.e., High-dimensional Sparse Cross-modal Hashing, HSCH for short. It leverages not only high-level semantic labels but also low-level multi-modal features to construct a fine-grained similarity. In particular, based on two well-designed rules, i.e., multi-level and prioritized, it is able to avoid semantic conflicts. Additionally, it leverages the strong power of high-dimensional sparse hash codes to preserve the fine-grained similarity. Then, it efficiently solves the sparse and discrete constraints of sparse hash codes through an efficient discrete optimization algorithm. In light of this, it is much more efficient and scalable to large-scale datasets. More importantly, the computational complexity of HSCH in the retrieval phase is as efficient as those naive hashing methods that use dense hash codes. Moreover, to support online learning scenarios, this paper also extends HSCH into an online version, i.e., HSCH_on. Extensive experiments on three benchmark datasets demonstrate the superiority of our framework compared with some state-of-the-art cross-modal hashing approaches in terms of both accuracy and efficiency.

Ensemble learning framework for image retrieval via deep hash ranking

Deep hashing combines feature extraction or representation with hash coding jointly, which can significantly improve the speed of large-scale image retrieval. However, we notice that compared with traditional retrieval methods, due to the reduction of dimension and information loss, the retrieval performance of binaryhash coding has declined to a certain extent. Most hash retrieval algorithms focus on the semantic similarity between image pairs, and ignore the ranking information between the returned samples. The returned samples should not only match the retrieved samples, but also rank the correct samples in front of the returned list. In addition, the performance difference of the deep model used in deep hash retrieval will also limit the efficiency of retrieval. To address such problem, we proposed an ensemble deep neural model robust framework for image retrieval, which can learn compact hash codes containing rich semantic information through hash constraints. The ensemble strategy is introduced, and the weighted voting is applied to integrate the ranking list. Comprehensive experiments on three benchmark datasets show that the proposed method achieves very competitive results. Codes are available at https://github.com/lidonggen-123/Ensemble_Deephash_Image_Retrieval.

High-resolution atmospheric retrievals of WASP-121b transmission spectroscopy with ESPRESSO: Consistent relative abundance constraints across multiple epochs and instruments

ABSTRACT Recent progress in high-resolution transmission spectroscopy has offered new avenues in which to characterize the atmospheres of transiting exoplanets. High-resolution cross-correlation spectroscopy allows for the unambiguous detection of molecules/atoms. It has also been used to map both atmospheric dynamics and longitudinal variations in the abundance of species across the morning and evening limbs. We present multiple Very Large Telescope (VLT)/Echelle SPectrograph for Rocky Exoplanets and Stable Spectroscopic Observations (ESPRESSO) observations of the ultrahot Jupiter WASP-121b, from which we constrain relative abundances of various neutral metals consistently across all observations, whilst accounting for the distortion of the exoplanet’s signal caused by traditional data processing techniques. We also constrain planetary orbital velocities and T–P profiles. We compare our abundance constraints with previous constraints using VLT/Ultraviolet and Visual Echelle Spectrograph (UVES) transmission spectroscopy of WASP-121b, and find our results to be consistent between observations, and also in agreement with stellar values for species previously detected in the atmosphere of WASP-121b. Our retrieval framework can also be used to identify potential exospheric species, resulting in extended absorption features beyond the transit equivalent Roche limit of WASP-121b (ReqRL ∼ 1.3 Rp). Hα, Fe ii, and Ca ii were found to extend to high altitudes (1.54 ± 0.04 Rp, 1.17 ± 0.01 Rp, and 2.52 ± 0.34 Rp, respectively), which are broadly consistent with literature values. The consistency of our constraints across multiple high-resolution observations is a strong validation of our model filtering and retrieval framework, as well as the stability of the atmosphere over the time-scales of months/years, and could allow for planet formation processes to be inferred from future ground-based observations of exoplanetary atmospheres.

Projected refractive index framework for multi-wavelength phase retrieval.

Multi-wavelength phase retrieval provides a competitive solution to lensless holographic imaging that features a low-cost, compact design and high data acquisition speed. However, the existence of phase wraps poses a unique challenge for iterative reconstruction, and the resulting algorithms often suffer from limited generalizability and increased computational complexity. Here, we propose a projected refractive index framework for multi-wavelength phase retrieval that directly recovers the amplitude and unwrapped phase of the object. General assumptions are linearized and integrated into the forward model. Based on an inverse problem formulation, physical constraints and sparsity priors are incorporated, which ensures imaging quality under noisy measurements. We experimentally demonstrate high-quality quantitative phase imaging on a lensless on-chip holographic imaging system using three color LEDs.

A study to develop a physiological framework for information retrieval using biological models manipulated by researchers

A study to develop a physiological framework for information retrieval using biological models manipulated by researchers

Universal Weighting Metric Learning for Cross-Modal Retrieval.

Cross-modal retrieval has recently attracted growing attention, which aims to match instances captured from different modalities. The performance of cross-modal retrieval methods heavily relies on the capability of metric learning to mine and weight the informative pairs. While various metric learning methods have been developed for unimodal retrieval tasks, the cross-modal retrieval tasks, however, have not been explored to its fullest extent. In this paper, we develop a universal weighting metric learning framework for cross-modal retrieval, which can effectively sample informative pairs and assign proper weight values to them based on their similarity scores so that different pairs favor different penalty strength. Based on this framework, we introduce two types of polynomial loss for cross-modal retrieval, self-similarity polynomial loss and relative-similarity polynomial loss. The former provides a polynomial function to associate the weight values with self-similarity scores, and the latter defines a polynomial function to associate the weight values with relative-similarity scores. Both self and relative-similarity polynomial loss can be freely applied to off-the-shelf methods and further improve their retrieval performance. Extensive experiments on two image-text retrieval datasets, three video-text retrieval datasets and one fine-grained image retrieval dataset demonstrate that our proposed method can achieve a noticeable boost in retrieval performance.

A Modified LBP Operator-Based Optimized Fuzzy Art Map Medical Image Retrieval System for Disease Diagnosis and Prediction.

Medical records generated in hospitals are treasures for academic research and future references. Medical Image Retrieval (MIR) Systems contribute significantly to locating the relevant records required for a particular diagnosis, analysis, and treatment. An efficient classifier and effective indexing technique are required for the storage and retrieval of medical images. In this paper, a retrieval framework is formulated by adopting a modified Local Binary Pattern feature (AvN-LBP) for indexing and an optimized Fuzzy Art Map (FAM) for classifying and searching medical images. The proposed indexing method extracts LBP considering information from neighborhood pixels and is robust to background noise. The FAM network is optimized using the Differential Evaluation (DE) algorithm (DEFAMNet) with a modified mutation operation to minimize the size of the network without compromising the classification accuracy. The performance of the proposed DEFAMNet is compared with that of other classifiers and descriptors; the classification accuracy of the proposed AvN-LBP operator with DEFAMNet is higher. The experimental results on three benchmark medical image datasets provide evidence that the proposed framework classifies the medical images faster and more efficiently with lesser computational cost.

Distributed search and fusion for wine label image retrieval.

With the popularity of wine culture and the development of artificial intelligence (AI) technology, wine label image retrieval becomes more and more important. Taking an wine label image as an input, the goal of this task is to return the wine information that the user hopes to know, such as the main brand and sub-brand of the wine. The main challenge in wine label image retrieval task is that there are a large number of wine brands with the imbalance of their sample images which strongly affects the training of the retrieval system based on deep learning. To solve this problem, this article adopts a distribted strategy and proposes two distributed retrieval frameworks. It is demonstrated by the experimental results on the large scale wine label dataset and the Oxford flowers dataset that both our proposed distributed retrieval frameworks are effective and even greatly outperform the previous state-of-the-art retrieval models.

A Clear View of a Cloudy Brown Dwarf Companion from High-resolution Spectroscopy

Direct imaging studies have mainly used low-resolution spectroscopy (R ∼ 20–100) to study the atmospheres of giant exoplanets and brown dwarf companions, but the presence of clouds has often led to degeneracies in the retrieved atmospheric abundances (e.g., carbon-to-oxygen ratio, metallicity). This precludes clear insights into the formation mechanisms of these companions. The Keck Planet Imager and Characterizer (KPIC) uses adaptive optics and single-mode fibers to transport light into NIRSPEC (R ∼ 35,000 in the K band), and aims to address these challenges with high-resolution spectroscopy. Using an atmospheric retrieval framework based on petitRADTRANS, we analyze the KPIC high-resolution spectrum (2.29–2.49 μm) and the archival low-resolution spectrum (1–2.2 μm) of the benchmark brown dwarf HD 4747 B (m = 67.2 ± 1.8 M Jup, a = 10.0 ± 0.2 au, T eff ≈ 1400 K). We find that our measured C/O and metallicity for the companion from the KPIC high-resolution spectrum agree with those of its host star within 1σ–2σ. The retrieved parameters from the K-band high-resolution spectrum are also independent of our choice of cloud model. In contrast, the retrieved parameters from the low-resolution spectrum are highly sensitive to our chosen cloud model. Finally, we detect CO, H2O, and CH4 (volume-mixing ratio of log(CH4) = −4.82 ± 0.23) in this L/T transition companion with the KPIC data. The relative molecular abundances allow us to constrain the degree of chemical disequilibrium in the atmosphere of HD 4747 B, and infer a vertical diffusion coefficient that is at the upper limit predicted from mixing length theory.

Sparse phase retrieval using a physics-informed neural network for Fourier ptychographic microscopy.

In this paper, we report a sparse phase retrieval framework for Fourier ptychographic microscopy using the recently proposed principle of physics-informed neural networks. The phase retrieval problem is cast as training bidirectional mappings from the measured image space with random noise and the object space to be reconstructed, in which the image formation physics and convolutional neural network are integrated. Meanwhile, we slightly modify the mean absolute error loss function considering the signal characteristics. Two datasets are used to validate this framework. The results indicate that the proposed framework is able to reconstruct sparsely sampled data using a small aperture overlapping rate without additional data driving whereas conventional methods cannot.

Using GNSS-IR Snow Depth Estimation to Monitor the 2022 Early February Snowstorm over Southern China

Snow depth is an essential meteorological indicator for monitoring snow disasters. The Global Navigation Satellite System Interferometric Reflectometry (GNSS-IR) technique has been proven to be a practical approach to retrieving snow depth. This study presents a case study to explore utilizing the GNSS-IR-derived snow depth to monitor the 2022 early February snowstorm over southern China. A snow depth retrieval framework considering data quality control and specific ground surface substances was developed using 8-day data from 13 operational GNSS/Meteorology stations. The daily snow depths retrieved from different ground surfaces, i.e., dry grass, wet grass, and concrete, agreed well with the measured snow depth, with Mean Absolute Error (MAE) of 2.79 cm, 3.36 cm, and 2.53 cm, respectively. The percentage MAE when snow depths > 5 cm for the three ground surface substances was 26.8%, 53.7%, and 35.0%, respectively. The 6 h snow depth results also showed a swift and significant response to the snowfall event. This study proves the potential of GNSS-IR, used as a new operational tool in the automatic meteorological system, to monitor snow disasters over southern China, particularly as an efficient and cost-effective framework for real-time and accurate monitoring.

SPARQL and SKOS Enabled Multilingual Polyhierarchical Thesaurus : Designing a Framework

The goal of this work is to investigate a multilingual thesaurus with a polyhierarchical structure based on TemaTres to publish vocabulary on the web for the simple retrieval of multilingual information by users. It could easily retrieve related metadata from other sources to improve advanced library services. The TemaTres and LAMP Architecture built and created this integrated prototype framework on the Ubuntu platform. Apart from that, it investigates Google Translate for easy data entry in various terms and facets across many languages. It must configure vocabulary settings to build a multilingual enabled polyhierarchical structure based on SPARQL Endpoint and API services. The study's primary outcomes include the development of a multilingual integrated framework for easy retrieval of information on various terms and facets by users. A multilingual data entry framework on a specific topic can be created using Unicode and TemaTres. There are numerous scripts and strategies available for web publishing. This study is unique in that it focuses on creating and retrieving multilingual thesaurus based on polyhierarchical relationships. The library professionals and users will benefit significantly from these integrated vocabulary services. It makes SPARQL linked data facilities by providing a visual vocabulary and thesaurus construction interface based on Web 2.0 and API services.