Automatic speech emotion recognition (SER) is a challenging task that plays a crucial role in natural human-computer interaction. One of the main challenges in SER is data scarcity, i.e., insufficient amounts of carefully labeled data to build and fully explore complex deep learning models for emotion classification. This paper aims to address this challenge using a transfer learning strategy combined with spectrogram augmentation. Specifically, we propose a transfer learning approach that leverages a pre-trained residual network (ResNet) model including a statistics pooling layer from speaker recognition trained using large amounts of speaker-labeled data. The statistics pooling layer enables the model to efficiently process variable-length input, thereby eliminating the need for sequence truncation which is commonly used in SER systems. In addition, we adopt a spectrogram augmentation technique to generate additional training data samples by applying random time-frequency masks to log-mel spectrograms to mitigate overfitting and improve the generalization of emotion recognition models. We evaluate the effectiveness of our proposed approach on the interactive emotional dyadic motion capture (IEMOCAP) dataset. Experimental results indicate that the transfer learning and spectrogram augmentation approaches improve the SER performance, and when combined achieve state-of-the-art results.

This paper presents a category theoretic ontology of Carnatic music. Our goals here are twofold. First, we will demonstrate the power and flexibility of conceptual modeling techniques based on a branch of mathematics called category theory (CT), using the structure of Carnatic music as an example. Second, we describe a platform for collaboration and research sharing in this area. The construction of this platform uses formal methods of CT (colimits) to merge our Carnatic ontology with a generic model of music information retrieval tasks. The latter model allows us to integrate multiple analytical methods, such as hidden Markov models, machine learning algorithms, and other data mining techniques like clustering, bagging, etc., in the analysis of a variety of different musical features. Furthermore, the framework facilitates the storage of musical performances based on the proposed ontology, making them available for additional analysis and integration. The proposed framework is extensible, allowing future work in the area of ra̅ga recognition to build on our results, thereby facilitating collaborative research. Generally speaking, the methods presented here are intended as an exemplar for designing collaborative frameworks supporting reproducibility of computational analysis and simulation.

BACKGROUND: Cell-scaffold contact measurements are derived from pairs of co-registered volumetric fluorescent confocal laser scanning microscopy (CLSM) images (z-stacks) of stained cells and three types of scaffolds (i.e., spun coat, large microfiber, and medium microfiber). Our analysis of the acquired terabyte-sized collection is motivated by the need to understand the nature of the shape dimensionality (1D vs 2D vs 3D) of cell-scaffold interactions relevant to tissue engineers that grow cells on biomaterial scaffolds. RESULTS: We designed five statistical and three geometrical contact models, and then down-selected them to one from each category using a validation approach based on physically orthogonal measurements to CLSM. The two selected models were applied to 414 z-stacks with three scaffold types and all contact results were visually verified. A planar geometrical model for the spun coat scaffold type was validated from atomic force microscopy images by computing surface roughness of 52.35 nm ±31.76 nm which was 2 to 8 times smaller than the CLSM resolution. A cylindrical model for fiber scaffolds was validated from multi-view 2D scanning electron microscopy (SEM) images. The fiber scaffold segmentation error was assessed by comparing fiber diameters from SEM and CLSM to be between 0.46% to 3.8% of the SEM reference values. For contact verification, we constructed a web-based visual verification system with 414 pairs of images with cells and their segmentation results, and with 4968 movies with animated cell, scaffold, and contact overlays. Based on visual verification by three experts, we report the accuracy of cell segmentation to be 96.4% with 94.3% precision, and the accuracy of cell-scaffold contact for a statistical model to be 62.6% with 76.7% precision and for a geometrical model to be 93.5% with 87.6% precision. CONCLUSIONS: The novelty of our approach lies in (1) representing cell-scaffold contact sites with statistical intensity and geometrical shape models, (2) designing a methodology for validating 3D geometrical contact models and (3) devising a mechanism for visual verification of hundreds of 3D measurements. The raw and processed data are publicly available from https://isg.nist.gov/deepzoomweb/data/ together with the web -based verification system.

This poster presents the problem of 3D contact measurements from two co-registered volumetric images (z-stacks). The 3D contact measurement consists of (a) segmenting an object of interest in each z-stack, (b) computing the relative spatial positions of the detected objects to detect contacts, (c) validating the accuracy of segmentation, and (d) visually verifying correct contact detection. The 3D measurement has to overcome challenges related to (1) intensity bleed-through across co-registered volumes, (2) insufficient knowledge about statistics and geometry of objects, (3) large RAM requirements (~3GB just to load the input data) and data volume (>1TB), and (4) complexity of 3D visual inspection.

Supporting different stakeholder viewpoints across the product lifecycle requires semantic richness to represent product-related information and enable multiview engineering simulations. This paper proposes a multilevel product modeling framework enabling stakeholders to define product models and relate them to physical or simulated instances. The framework is defined within the Model-Driven Architecture using the multilevel (data, model, metamodel) approach. The data level represents real-world products, the model level describes models (product models) of real-world products, and the metamodel level describes models of the product models. The metamodel defined in this paper is specialized from a web ontology language enabling product designers to express the semantics of product models in an engineering-friendly way. The interactions between these three levels are described to show how each level in the framework is used in a product engineering context. A product design scenario and user interface for the product metamodel is provided for further understanding of the framework.

Tolerances impose restrictions on the possible deviations of features from their nominal sizes/shapes. These variations of size/shape could be thought of as deviations of a set of generalized coordinates defined at some convenient point on a feature. Any tolerance specification for a feature imposes some kind of restrictions or constraints on its deviation parameters. These constraints, in general, define a bounded region in the deviation space. In this paper, a method has been presented for converting tolerance specifications as per maximum material condition (MMC)/least material condition (LMC)/regardless of feature size (RFS) material conditions for standard mating features (planar, cylindrical, and spherical) into a set of inequalities in a deviation space. Both virtual condition boundaries Virtual condition boundary: A constant boundary generated by the collective effects of a size feature's specified MMC or LMC material condition and the geometric tolerance for that material condition. (VCB) and tolerance zones are utilized for these mappings. The mapping procedures have been illustrated with an example. Note to Practitioners-This paper deals with methods to convert tolerance specification as per ASME Y14.5M into a set of generalized deviation of features and vice-versa. These are intermediate relationships that are required for use in deviation-based tolerance synthesis methods. In this work, different examples have been presented to show how different tolerance specifications (such as positional tolerance at maximum material condition (MMC), least material condition (LMC), etc.) applied to different features could be treated on a generalized basis for tolerancing of manufactured parts.

This article describes the design and development of a design repository software system. This system is a prototype implementation intended to demonstrate the role of design repositories as part of a vision for the next generation of product development software systems. This research involves not only the creation of a prototype software system, but is part of a broader effort that also includes the development of a core product knowledge representation and that seeks to address terminological and semantic issues associated with computer-aided product development. This article focuses on the interfaces that have been developed to support authoring and navigation of the product models stored in design repositories as well as the software architecture and associated rationale that provide the framework on which the system is built.

An increasing trend toward product development in a collaborative environment has resulted in the use of various software tools to enhance the product design. This requires a meaningful representation and exchange of product data semantics across different application domains. This paper proposes an ontology-based framework to enable such semantic interoperability. A standards-based approach is used to develop a Product Semantic Representation Language (PSRL). Formal description logic (DAML+OIL) is used to encode the PSRL. Mathematical logic and corresponding reasoning is used to determine semantic equivalences between an application ontology and the PSRL. The semantic equivalence matrix enables resolution of ambiguities created due to differences in syntaxes and meanings associated with terminologies in different application domains. Successful semantic interoperability will form the basis of seamless communication and thereby enable better integration of product development systems. Note to Practitioners-Semantic interoperability of product information refers to automating the exchange of meaning associated with the data, among information resources throughout the product development. This research is motivated by the problems in enabling such semantic interoperability. First, product information is formalized into an explicit, extensible, and comprehensive product semantics representation language (PSRL). The PSRL is open and based on standard W3C constructs. Next, in order to enable semantic translation, the paper describes a procedure to semi-automatically determine mappings between exactly equivalent concepts across representations of the interacting applications. The paper demonstrates that this approach to translation is feasible, but it has not yet been implemented commercially. Current limitations and the directions for further research are discussed. Future research addresses the determination of semantic similarities (not exact equivalences) between the interacting information resources.

In this paper, a generic model for the synthesis of tolerances for manufactured parts is presented. The model uses a method of transforming traditional tolerance specifications (as defined in ASME Y14.5M) to a generalized coordinate system (hereinafter referred to as deviation space). Small displacement torsors (SDTs) have been used for representing the deviations. The tolerance synthesis method is formulated as a constrained nonlinear optimization process. Three different types of constraints have been considered for the optimization process: 1) representation of assemblability; 2) mapping of tolerance specification to deviations; and 3) functional requirements. A new deviation-based cost of manufacturing model has been proposed. A working module of the scheme has been implemented and the process has been elaborated with two examples. The possibility of extension of the model and scope for further generalization have been discussed. Note to Practitioners-This paper presents an optimization method for finding tolerance values for different features of an assembly of manufactured parts. Determination of different types of tolerances and their exact values for critical features of any part has been an ad-hoc process; it has been mostly experience-based till recent time. In this paper, efforts have been made to establish a strong mathematically oriented method for synthesis of tolerances. The procedure attempts to minimize cost of manufacturing while the functionality and assemblability are satisfied. It is a generalized method and could be applied for designing rigid manufacture parts. For practical usage, this method should be integrated with three-dimensional computer-aided design packages as a tolerance synthesis module for integrated tolerance design.

Tolerancing decisions can profoundly impact the quality and cost of electro-mechanical assemblies. Existing approaches to tolerance analysis and synthesis in design entail detailed knowledge of geometry of the assemblies and are mostly applicable during advanced stages of design, leading to a less than optimal design process. During the design process of assemblies, both the assembly structure and associated tolerance information evolve continuously. Therefore, significant gains can be achieved by effectively using this information to influence the design of the assembly. Motivated by this, we identify and explore two goals for future research that we believe can enhance the scope of tolerancing for the entire design process. The first goal is to advance tolerancing decisions to the earliest possible stages of design. This issue raises the need for effective representation of tolerancing information during different stages of design and for effective assembly modeling. The second goal addresses the appropriate, synergistic use of available methods and best practices for tolerance analysis and synthesis, at successive stages of design. Pursuit of these goals leads to the definition of a multilevel approach that enables tolerancing to be addressed at successive stages of design in an incremental fashion. The resulting design process, which we call the design for tolerance process, integrates three important domains: (1) design activities at successive stages of design; (2) assembly models that evolve continuously through the design process; (3) methods and best practices for tolerance analysis and synthesis. We demonstrate major steps of our proposed approach through a simple, yet illustrative, example.

Tolerancing decisions during the design of electromechanical products profoundly affect cost and quality. Existing approaches to tolerance analysis and synthesis entail detailed knowledge of geometry of the assemblies and are mostly applicable during advanced stages of design, leading to a less than optimal design process. During design, both the assembly structure and associated tolerance information evolve continuously and significant gains can be achieved by effectively using this information to influence the design. We explore two goals of research to expand the scope of tolerancing to the entire design process. The first goal is to advance tolerancing decisions to the earliest possible stages of design. This issue raises the need for effective representation of tolerancing information during early stages. The second goal addresses the appropriate use of industry best practices and efficient computational approaches. Pursuit of these goals leads to the definition of a multilevel approach that enables tolerancing to be addressed at successive stages of design in an incremental, continuous ongoing fashion. The resulting design process, which we call the design-for-tolerance process, integrates three important domains: (1) design activities at successive stages of design; (2) assembly models that evolve continuously through the design process; and (3) methods and best practices for tolerance analysis and synthesis. We demonstrate major steps of our proposed approach through a simple, yet illustrative, example.

We report on the development of a knowledge representation model, which is based on the SHARED object model reported in Shared Workspaces for Computer-Aided Collaborative Engineering (Wong, A. and Sriram, D., Technical Report, IESL 93-06, Intelligent Engineering Systems Laboratory, Department of Civil Engineering, MIT, March, 1993) and Research in Engineering Design (Wong, A. and Sriram, D., SHARED: An Information Model for Cooperative Product Development, 1993, Fall, 21-39). Our current model is implemented as a layered scheme, that incorporates both an evolving artifact and its associated design process. To represent artifacts as they evolve, we define objects recursively without a pre-defined granularity on this recursive decomposition. This eliminates the need for translations between levels of abstraction in the design process. The SHARED model extends traditional OOP in three ways: by allowing explicit relationship classes with inheritance hierarchies; by permitting constraints to be associated with objects and relationships; and by comparing `similar' objects at three different levels (form, function and behavior).