Samsung Developer Conference 2024 (SDC24) Coming Up
Samsung Developer Conference 2024 (SDC24) is coming! This global event for developers, hosted by Samsung Electronics, will take place on October 3, at the San Jose McEnery Convention Center, California, USA. Celebrating its 10th year, SDC24 will offer sessions on diverse topics and interesting exhibitions, as well as the keynote presentation which is the pinnacle of the event. Various programs for attendees are also in the works including meetups, Code Labs, and dinner parties, so please stay tuned! SDC24 is open to everyone, free of charge. For more about the event, such as participant registration and an exclusive hotel reservation page for SDC24, see the links below.
|
|
|
|
Non-Disruptive Cloud Infrastructure Upgrade Architecture for Samsung Account
As an account service that connects over 60 services and applications in 256 countries around the world, Samsung Account is a core element of services offered by Samsung Electronics. Samsung Account is used for authentication on mobile, wearable, TV, PC, and more, and enhances customer experiences at online stores like Samsung.com, as well as offline stores.
The infrastructure and services of Samsung Account have continuously evolved and improved, and it is currently serving 4 regions, including 3 global regions and China. More than 70 microservices make up Samsung Account, which is based on AWS EKS.
However, managed services such as AWS EKS, while convenient, have the drawback of needing to be upgraded when their support ends. Samsung Account has adopted multi-cluster architecture for this purpose, and has successfully upgraded EKS using this architecture. Learn about the architecture Samsung Account used to carry out the non-disruptive EKS upgrade under high traffic.
|
|
|
|
Samsung IAP Subscription Got More Robust with New Features!
We have released a new real-time notification service for developers and open API for subscriptions. Manage your customers and sales more easily and efficiently.
Real-time developer notifications
Try out the new real-time developer notifications. You can add a URL to receive notifications on purchases or refunds of in-app items, or for subscription status changes of your subscribers. You can also retrieve refunded items or send personalized suggestions to individual users according to their status changes such as unsubscribing, managing them in a customized way.
Learn more
Open API for subscriptions
Developers can use "Galaxy Store Developer API" when they need to directly control their subscription product's cancellation, refund, termination, and more. They can also provide options for their subscribers to carry out these operations through their application or website themselves.
|
|
|
|
Tutorial: Craft Adaptive UI for Android Applications Using Jetpack Compose
Due to the large variety of Android devices, supporting multiple screen sizes and display modes is a necessity in the modern Android application development. Using Android's new adaptive library, you can change the layout of your application to support foldable and large screen devices in portrait, landscape, split-screen and pop-up modes. Learn how to create a consistent and optimized user experience across various screen sizes and devices in this tutorial.
|
|
|
|
Introducing the Works with SmartThings (WWST) Test Suite
At SmartThings, we’re committed to providing developers with the tools they need to create innovative and seamless smart home experiences. We are excited to introduce our newest addition to the developer toolkit: the Works with SmartThings (WWST) Test Suite.
The Test Suite is designed to assist our partners in certifying their devices as part of the WWST certification process. With Test Suite, developers can easily test SmartThings integrations, get certified faster, and showcase their devices to tens of millions of SmartThings users. Click the link below to learn more.
|
|
|
|
UPR-Net: Unified Pyramid Recurrent Network for Video Frame Interpolation
Video frame interpolation (VFI) is a classic low-level vision task that synthesizes non-existent intermediate frames between originally consecutive frames, playing an essential role in many practical applications including novel view synthesis, video compression, cartoon creation, and so on. The existing technique of flow-guided synthesis is a popular framework for video frame interpolation, where optical flow is estimated to warp input frames first, and then the intermediate frame is synthesized from warped representations. However, it is common that the intermediate frame is synthesized in a single pass, possibly missing the opportunity of refining imperfect synthesis for high-resolution and large motion cases.
This article introduces a new technique for frame interpolation called Unified Pyramid Recurrent Network (UPR-Net). Cast in a flexible pyramid framework, UPR-Net exploits lightweight recurrent modules for both bi-directional flow estimation and intermediate frame synthesis. UPR-Net leverages the estimated bi-directional flow to generate forward-warped representations for frame synthesis at each pyramid level, and enables iterative refinement for both the optical flow and intermediate frame, significantly improving the interpolation robustness on large motion cases. UPR-Net trained on low-resolution data can achieve excellent performance on both low- and high-resolution benchmarks, as well as being able to compete favorably with many methods that rely on much heavier architectures, despite being extremely lightweight. Learn more about UPR-Net on the Samsung Research blog.
Learn more |
|
|
|
Object-Centric Representation Learning for Video Scene Understanding
Video understanding is crucial for various real-world applications, such as surveillance, autonomous driving, augmented reality, and the metaverse. Depth-aware Video Panoptic Segmentation (DVPS) is a task designed to achieve this goal of video understanding. In general, all panoptic objects in a video can be categorized into foreground instances (things), such as people, cars, etc., and background semantics (stuff), such as the sky, road, etc. The DVPS task requires estimating the semantic class and 3D depth for all things and stuff and then uniquely segmenting and consistently tracking all items in the scene. Prior methods typically approach the DVPS task as a multi-task learning problem, employing a multi-branch framework, and handle things and stuff separately in a video. However, this approach requires intricate post-processing and falls short in fully harnessing the relationships between different tasks.
We have introduced a unified object-centric pipeline called Slot-IVPS to effectively leverage the relationship between depth and semantic information. This technique, integrating semantic and depth information for all panoptic objects within a video, is the first fully unified end-to-end pipeline for the DVPS task. It fully leverages the semantic and depth information without requiring multiple separate representations or task-specific branches. The pipeline is validated on both DVPS and VPS tasks, its performance surpassing the latest methods. Learn more about this on the Samsung Research blog.
Learn more |
|
|
|
Uniting Explainable and Connectionist Approaches for Visual Question Answering
Visual Question Answering (VQA) is a field of research that aims to develop methods for answering natural language questions based on the information provided in given images. It is being applied in many different areas including human-machine interactions, robotics, education, healthcare, the metaverse, etc. Most VQA methods learn to answer questions based on the statistical correlations between multimodal inputs and their answers. In order to address the issue of superficial correlations between questions and answers, recent studies represent the visual information as a structured scene graph or convert the questions into a program of executable neural modules.
In this article, we propose a new method where we join explainable and connectionist modules for answering questions. This approach offers robust visual reasoning and question answering with a novel architecture that leverages question-guided attention to combine explainable and connectionist approaches. The VQA performance is enhanced through the training and learning of the contributions of explainable and connectionist networks. Check out the Samsung Research blog to learn about this novel and high-performance approach.
Learn more |
|
|
|
|
|
|
|
https://developer.samsung.com
Copyright© %%xtyear%% SAMSUNG All Rights Reserved.
This email was sent to %%emailaddr%% by Samsung Electronics Co.,Ltd.
You are receiving this email because you have subscribed to the Samsung Developer Newsletter through the website.
Samsung Electronics · 129 Samsung-ro · Yeongtong-gu · Suwon-si, Gyeonggi-do 16677 · South Korea
Privacy Policy
Unsubscribe
|
|
|
|
|
|
|
|
|
