Faculty Awards

Boston University Computer Science Faculty Win 3 National Science Foundation CAREER Awards

In May 2026, three members of Boston University’s Department of Computer Science Faculty earned National Science Foundation Career Awards to support their research projects.

By: dannie
May 29, 2026

Boqing Gong

Assistant Professor in Computer Science Dr. Boqing Gong received a 5-year grant for $599,999 in support of his project, “Democratizing the Pretraining of Vision Foundation Models: A Developmentally Plausible Framework.”

Here is an abstract of the project:
Vision foundation models (VFMs) are artificial intelligence systems for “all-purpose” understanding of images and videos. They are currently extremely expensive to create. This high cost restricts their creation to a few highly resourced institutions and leaves independent researchers and the public unable to fully explore how these systems learn. This project seeks to democratize this research by creating a highly efficient training method inspired by how human infants learn. A human child acquires foundational visual skills from a limited number of waking hours compared to the massive amount of data used by current VFMs. By using longitudinal video and audio recorded from the viewpoint of infants, this project develops a training process that is affordable for university budgets. Innovating and understanding how to train these systems efficiently using this infant-inspired approach will increase accessibility to artificial intelligence research for the broader public. Furthermore, the project provides unique educational opportunities for students and offers insights that can be transferred to specialized industries, such as medical imaging and vocational training, where data is often limited. Expanding community involvement in building these models will ultimately promote artificial intelligence safety, enhance transparency, and build public trust.

The technical goal of this project is to formalize a developmentally plausible, data-efficient pretraining framework for VFMs. First, the team of researchers will establish a core framework by curating longitudinal, egocentric audiovisual recordings of human infants and designing a suite of evaluation benchmarks strictly aligned with early cognitive milestones. Second, the project bridges inherent sensory and temporal gaps in the recordings. This involves employing model ensembling to simulate tactile and gustatory senses from audiovisual cues and utilizing a meta-learning formulation to optimally mix heterogeneous data sources. Third, the investigators will design novel model architectures and pretraining algorithms tailored for a continuous “baby learning” paradigm. To achieve this, the research incorporates continuous-state Hopfield networks to serve as an expansive associative memory module, which mitigates catastrophic forgetting. Moreover, the project introduces a monotonic neural network for non-linear uncertainty calibration without sacrificing the accuracy of the pretext tasks. By integrating these three thrusts, the project will yield open-source baseline models, developmental benchmarks, and algorithms that enable the broader scientific community to investigate highly efficient pretraining methodologies.

Assistant Professor in Computer Science Dr. John Liagouris received a 5-year grant for $680,519 in support of his project, “A Unified Analytics Stack for Secure Computation.”

Here is an abstract of the project:
This project introduces a unified software stack for secure computation that integrates cryptographic and hardware-based techniques, each with its own unique strengths, challenges, and performance characteristics. The project’s novelties include (i) software abstractions and intermediate representations that allow reusing functionality across technologies and workloads, (ii) a distributed and fault-tolerant system runtime for secure data analysis pipelines, and (iii) a versatile performance modeling and optimization framework that integrates diverse cost metrics to efficiently deploy secure data pipelines in heterogeneous environments. The project’s broader significance is the potential to enable secure analytics in a scalable fashion; an ability that will have implications on how modern society protects privacy and intellectual property while extracting value from data.

The project includes three complementary thrusts that focus on software abstractions, scalable workload distribution, and cost-based optimization. The project designs a unified software architecture for secure analytics that supports diverse technologies (fully homomorphic encryption, secure multiparty computation, trusted execution environments) and workloads (machine learning, relational analytics, time series computations) on top of the same oblivious execution engine. Second, the project develops a novel parallel and distributed system runtime that scales secure computation within and across machines, leveraging heterogeneous resources and ensuring transparent fault tolerance. Finally, the project introduces original cost-based optimization techniques that incorporate performance objectives, threat models, and monetary budgets to enable automated planning of secure data pipelines. The project aims to make privacy-enhancing technologies a core component of the computer science education and to lay the foundation for a new generation of secure computing systems by rethinking the entire analytics stack: from the programming abstractions all the way down to the hardware.

Assistant Professor in Computer Science Dr. Ed Chien received a 5-year grant for $649,000 in support of his project, “Designing Vector Field Flows for Computational Knitting and Curved Layer 3D Printing.”

Here is an abstract of the project:
Computational fabrication via technologies such as 3D printing and computational knitting are key methods that form a significant part of modern advanced manufacturing, allowing for production of state-of-the-art composites, ceramics, medical grafts, and architectural formworks in complex geometries. Furthermore, recent advances have allowed for curved layer fabrication, producing objects and materials that have superior strength and quality characteristics due to control over build direction. Underlying many of these technologies is the fundamental problem of constructing a surface or volume from a single continuous curve, representing a toolpath or fiber path. To maximize utilization of these technologies, the research team will produce mathematical design frameworks for solving this problem under various fabrication modalities. The frameworks will be tailored to achieve domain-specific performance goals and accommodate domain-specific user design constraints. All resulting tools will be released as open-source implementations for use and further development by industry and academic researchers. Parts of the research will also be incorporated into coursework on geometry processing and graphics, and into graduate- and undergraduate-level research projects, via theses and summer research programs.

The research effort will be divided into three thrusts. First, the team will build upon prior work understanding the global topology of vector field flows on surfaces and construct an appropriate discretization and optimization framework that achieves the necessary path continuity and spacing constraints crucial to the fabrication modalities at hand. Second, the work will more closely explore application of the general optimization framework to the specific use case of computational knitting, where the space-filling curve follows the path of stitches that are produced by the machine. In this setting, the team will explore optimal geometric shaping, and incorporation of user design constraints as communicated by industry partners, who are using these methods to produce garments and curved surface composites. Thirdly, the team will look to extend the topological understanding and optimization frameworks to the challenging volumetric setting. This will target the nascent fabrication methodology of curved-layer 3D printing. The global topologies for volumetric fields are much more complex, and not yet fully understood from the theoretical perspective. Incorporation of structural and manufacturing constraints into layer design will also be considered, in collaboration with engineering colleagues.