# HCI Published articles for HCI. This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## \[GSoC 2026\] Bluetooth: HCI Improvements & HID Profile | Haiku Project DevFeed: [\[GSoC 2026\] Bluetooth: HCI Improvements & HID Profile | Haiku Project]() Original publisher: [Read original article]() Author: mohammedrattia Published: 2026-05-03T13:46:03Z Content type: article Language: en Sources: [Haiku Project]() Topics: [Bluetooth](), [Computer science](), [Development](), [Software Engineering](), [Operating system]() Tags: [2026](), [bluetooth](), [computer-science](), [google-summer-of-code](), [gsoc](), [gsoc2026](), [haiku](), [hci](), [hid](), [improvements](), [profile](), [programming](), [project](), [software](), [software-engineer](), [software-engineering](), [systems]() ### AI overview A GSoC 2026 proposal for Haiku aims to modernize the operating system's Bluetooth stack and implement the Bluetooth HID profile. The planned work includes completing HCI commands, improving pairing and Link Key persistence, and handling disconnection and timeout events. ### Source excerpt Hello! I'm Mohammed R. Attia, a 2nd-year Computer Science Student. I've been accepted into Google Summer of Code (GSoC) 2026 with Haiku, with my proposal titled "Modernizing Haiku's Bluetooth Stack: HCI Completion and HID Profile Implementation." About Me I consider myself a competitive programmer and a software engineer. I haven't decided on a specific specialization in software engineering just yet; rather, I'm interested in systems development and low-level programming. You can check my GitHub for more information about my latest projects. I go by the handle mohammedrattia in many places on the internet, so feel free to reach out :). How did I get here? I first learned about Haiku while searching for GSoC organizations to contribute to. It was my first time learning about a desktop operating system other than the famous trio: Windows, macOS, and Linux. So, I started by reading the GSoC guide and joining IRC. I remember that the process was smooth because the guide contained everything, and people on IRC were welcoming and supportive. Then, I started working towards building Haiku, and I got my first commit merged into Haiku after several days. Current Goals The main goal of my proposal in GSoC is to modernize the Bluetooth stack. Currently, the Bluetooth stack in Haiku can't communicate effectively with modern Bluetooth devices because of the outdated standards it uses. However, making the Bluetooth core perfectly align with modern standards won't benefit Haiku users because it would still need a Bluetooth Profile to perform a useful task. One of the most common uses of Bluetooth is Human Interface Devices (HID) (e.g., keyboards and mice). Accordingly, this project will build the HID Bluetooth profile to provide useful functionality for Haiku users. The following three sub-sections contain a brief specification for the goals of the project from my GSoC proposal. Modern HCI Event & Command Handling Build upon the recent SSP/EIR groundwork by implementing the remai ## Espressif says ESP32 undocumented Bluetooth HCI commands are not a backdoor DevFeed: [Espressif says ESP32 undocumented Bluetooth HCI commands are not a backdoor]() Original publisher: [Read original article]() Author: John Lee Published: 2025-03-10T00:00:00Z Content type: article Language: en Sources: [Blog on Developer Portal]() Topics: [ESP32](), [Bluetooth](), [Security](), [backdoor]() Tags: [backdoor](), [blog](), [bluetooth](), [esp32](), [hci](), [iot](), [security]() ### AI overview Espressif explains that the reported undocumented ESP32 Bluetooth HCI commands are controller debug commands. It says they do not participate in normal HCI communication with NimBLE or Bluedroid and do not pose a security threat or backdoor. ### Source excerpt Overview# Espressif has already provided a formal response to the recently published claims about ESP32 Bluetooth controller serving as a potential "backdoor" or having "undocumented features" that can cause security concerns. This post highlights the technical details about the relevant commands (HCI Commands) that were undocumented and establishes that the mentioned undocumented HCI commands do not pose a security threat and are certainly not a "backdoor". ## ScreenAI: A visual language model for UI and visually-situated language understanding DevFeed: [ScreenAI: A visual language model for UI and visually-situated language understanding]() Original publisher: [Read original article]() Author: Google AI (noreply@blogger.com) Published: 2024-03-19T20:15:00Z Content type: release Language: en Sources: [Google Research]() Topics: [vlm](), [ui](), [Google](), [datasets](), [multimodal](), [Transformer]() Tags: [datasets](), [google](), [hci](), [model](), [multimodal](), [multimodal-learning](), [self-supervised-learning](), [ui](), [vision]() ### AI overview Google Research introduces ScreenAI, a vision-language model designed to understand, reason about, and interact with user interfaces and infographics. The article describes its PaLI-based multimodal architecture, pix2struct-inspired flexible patching strategy, training approach, reported benchmark results, and three released datasets for screen understanding and question answering. ### Source excerpt Posted by Srinivas Sunkara and Gilles Baechler, Software Engineers, Google Research Screen user interfaces (UIs) and infographics, such as charts, diagrams and tables, play important roles in human communication and human-machine interaction as they facilitate rich and interactive user experiences. UIs and infographics share similar design principles and visual language (e.g., icons and layouts), that offer an opportunity to build a single model that can understand, reason, and interact with these interfaces. However, because of their complexity and varied presentation formats, infographics and UIs present a unique modeling challenge. To that end, we introduce "ScreenAI: A Vision-Language Model for UI and Infographics Understanding". ScreenAI improves upon the PaLI architecture with the flexible patching strategy from pix2struct. We train ScreenAI on a unique mixture of datasets and tasks, including a novel Screen Annotation task that requires the model to identify UI element information (i.e., type, location and description) on a screen. These text annotations provide large language models (LLMs) with screen descriptions, enabling them to automatically generate question-answering (QA), UI navigation, and summarization training datasets at scale. At only 5B parameters, ScreenAI achieves state-of-the-art results on UI- and infographic-based tasks (WebSRC and MoTIF), and best-in-class performance on Chart QA, DocVQA, and InfographicVQA compared to models of similar size. We are also releasing three new datasets: Screen Annotation to evaluate the layout understanding capability of the model, as well as ScreenQA Short and Complex ScreenQA for a more comprehensive evaluation of its QA capability. ScreenAI ScreenAI's architecture is based on PaLI, composed of a multimodal encoder block and an autoregressive decoder. The PaLI encoder uses a vision transformer (ViT) that creates image embeddings and a multimodal encoder that takes the concatenation of the image and text emb