Dev Insight May 2025: New Application Promotion Feature to Highlight Your Application in Galaxy Store, Meet the Samsung Galaxy S25 Edge, and Other Latest News

[Samsung Newsroom]

May 2025

New Application Promotion Feature to Highlight Your Application in Galaxy Store Introducing a new opportunity to increase your application’s visibility on the Galaxy Store! The Seller Portal has launched a new application promotion feature to highlight your application on the Galaxy Store. Developers can now request that their game, application, or theme is showcased in an editorial created for the Galaxy Store’s "Discover" tab. Available to users in the US and South Korea, the "Discover" tab is where discovery meets inspiration, highlighting the best content and original stories on mobile and tablet. The tab’s editorials provide users with curated recommendations on the latest trends, benefits and features exclusive to Galaxy, seasonal events, editors’ tips, and more, combining storytelling and visuals to enrich the user experience and showcase innovation. Check out our Seller Portal to learn more about having your application appear in an editorial and what you need to apply. Learn More Integrate Your Devices Easily with Device Profile Builder SmartThings continues to invest in the Developer Center by creating new tools for partners and has recently added the Device Profile Builder, a web-based tool that helps developers create a profile to define a device and its features on the SmartThings platform. The tool makes designing your device for SmartThings easier than ever. There are templates for different smart home device types, making it easy to get started, along with opportunities for customization. To learn more about the Device Profile Builder, see our blog article on the new tool. Meet the Samsung Galaxy S25 Edge: an Engineering Marvel of New Slim Hardware Innovation On May 13, Samsung unveiled their latest ultra-slim smartphone at the "Galaxy S25 Edge: Beyond Slim" event. At just 5.8 mm thin and weighing only 163 g, the Galaxy S25 Edge is the slimmest device in the Galaxy S series to date, combining sleek style with portability. The device features a titanium frame as well as Corning® Gorilla® Glass Ceramic 2 on the front display, a next-generation material designed to boost durability. Equipped with a 200-megapixel camera and enhanced AI capabilities, the Galaxy S25 Edge delivers an even more intelligent and powerful experience. Learn more about the Galaxy S25 Edge and its spectacular performance at the Samsung Electronics Newsroom. Learn More Tutorial: Reading and Converting Galaxy Watch Accelerometer Data Accessing accelerometer data from your Galaxy Watch opens up endless opportunities to build motion-aware applications—from fitness tools to gesture-controlled interfaces. With the proper setup and data handling, an application running on Galaxy Watch powered by Wear OS can deliver accurate, real-time motion insights to users. This tutorial walks you through everything you need to know, from setting up your Wear OS project to reading and converting sensor data. Explore the sample application included with the guide to test motion features yourself! Learn More Exploring a Simple Siamese Network for High-Resolution Video Quality Assessment Video Quality Assessment (VQA) is becoming increasingly important as popular smart devices make producing high-quality video and high-resolution user-generated content (UGC) part of everyday life. However, the computational burden for high-resolution video and the diversity of content present new challenges for efficient and effective VQA. To address these challenges, Samsung R&D Institute China-Nanjing presents SiamVQA, a Siamese network that considers both technical and aesthetic quality perspectives. Sharing a common feature extractor (Swin-T), SiamVQA employs a simple network design based on dual cross-attention to enhance the semantic perception ability for more accurate quality prediction. As a result, SiamVQA has achieved state-of-the-art accuracy on high-resolution benchmarks, proving to be both efficient and effective. Learn more about the next-generation VQA model that opens up the possibilities for Siamese networks on the Samsung Research blog. Learn More On-Device Hand Model for Robust Gesture Detection Hand gesture recognition technology enables more natural and intuitive ways to communicate. Unlike traditional input methods like keyboards or touchscreens, users can express intentions and commands through everyday movements and postures. Gesture-based interaction is also inclusive, offering greater accessibility for people with physical disabilities who may have difficulty using conventional input methods. Thanks to these advantages, hand gesture recognition is becoming an increasingly important area of research in different fields such as human-computer interaction systems, sign language translation, VR/AR, and remote-control applications.  In this blog article, we introduce an on-device AI model that delivers fast and accurate performance across a range of environments. Explore this key solution from Samsung R&D Institute Ukraine, from a neural network architecture to resolve multimodal tasks to a novel multi-cascade structure for handling simultaneous tasks at once. Learn More Single-Channel Distance-Based Source Separation for Mobile GPU in Outdoor and Indoor Environments Distance-based source separation (DSS) separates audio sources into "near" or "far" groups based on their distance from a microphone. Unlike traditional source separation approaches, DSS uses spatial cues—such as the inverse square law of sound intensity, direct-to-reverberation ratio (DRR), and other distance-based acoustic effects—to separate a mixed audio signal into components by distance. However, most previous DSS research has focused on indoor scenarios. This makes it difficult to address the challenges of ambient noise and unclear reverberant boundaries in outdoor environments, and of real-time processing speed and accuracy on mobile devices. To tackle this difficulty, AI Solution Team at Samsung Research proposes a new single-channel DSS model for on-device mobile environments. This model performs robustly in both indoor and outdoor scenarios. Key innovations of the model include a two-stage Conformer architecture for capturing complex audio dependencies, a linear relation-aware self-attention (RSA) mechanism for efficiency, and mobile GPU acceleration optimization using the TensorFlow Lite (TFLite) GPU delegate, which enables real-time running on mobile devices. Learn about how this new model achieves accurate and effective far/near voice and sound separation even in noisy outdoor environments and its on-device performance on the Samsung Research blog. Learn More

View the full blog at its source