🦄 Making great presentations more accessible.
This project aims to enhances multilingual accessibility and discoverability while maintaining the integrity of original content. Detailed transcriptions and keyframes preserve the nuances and technical insights that make each session compelling.
Overview
📖 AWS re:Invent 2025 - StradVision's Vision AI journey with AWS (GBL203)
In this video, Philp Kim and Insu Kim from StradVision present their session in Korean at the Global Impact Pavilion Theater 5. The presentation focuses on StradVision's Vision AI solutions and their data infrastructure on AWS. They discuss their migration of over 1 petabyte of data to AWS S3, data ingestion and processing pipelines, and challenges with traditional data transfer methods. The speakers cover their use of EKS/ECS for data processing and labeling workflows, GPU-based model training infrastructure, and various scenarios for data labeling processes in the cloud.
; This article is entirely auto-generated while preserving the original presentation content as much as possible. Please note that there may be typos or inaccuracies.
Main Part
StradVision Presentation on Vision AI and AWS Data Infrastructure (Korean Language Session)
Alright folks, we're going to get started here with our next section. This is our second to last session here at the Global Impact Pavilion Theater 5, and this one is a special one as mentioned, as it will be given and presented in Korean. So for those that speak Korean, you're in for a treat. Without further ado, I'll welcome to the stage Philp Kim and Insu Kim from StradVision.
Hello everyone. As you know, it's going to be in Korean, but you can still read the slides.
The transcript provided is severely corrupted and contains predominantly unintelligible text that does not correspond to the presentation content about StradVision's Vision AI solutions and AWS data infrastructure. The original speech-to-text output appears to have failed to accurately capture the speaker's words, resulting in a string of phonetically garbled syllables and fragments that cannot be reliably reconstructed into coherent sentences about data ingestion, processing pipelines, or the traditional data transfer challenges outlined in the referenced slides.
While the image tags marking positions 320, 330, 350, 410, 420, 430, 440, 450, 460, 500, 520, 540, 550, and 590 are preserved, the actual spoken content between and around these markers is too corrupted to produce a meaningful revision. The few recognizable phrases such as "data flow," "ingestion processing," "productivity scalability," and "international shipping" appear sporadically but lack sufficient context or surrounding coherent text to form complete thoughts or logical paragraphs.
To properly revise this transcript, a higher quality audio recording or an alternative transcription would be necessary, as the current source material does not contain enough recoverable information to reconstruct the presenter's intended message about StradVision's migration of over 1 petabyte of data to AWS S3, their use of EKS/ECS for data processing and labeling, or their GPU-based model training infrastructure.
You've got to engage with data connected to geo data transport terminal service. So he is monitoring the zone, then checking the data, then gas sim archiving service. So the archiving, she can push your transport officer to ensure the process. Process. So you've got data processing and competing. The western technology is all pushing on quality integration and automation. There's a scenario of your low labor, and then you've got to test the data. Computing premises. So engineer, yes. The nation is all on premises in pipeline, the cloud. The legal name flex. You look at packaging, the labeling to get the data. The labeling and the frame labeling, parking scenario, there's a labeling to the labeling. Processing and the kit is the labeling.
To give all, to go cloud frontal labeling of cloud frontal. The data labeling is a process that we need to handle carefully. Pushing in cloud print with data. We have various singing and data points. Sta flow in with various scenarios and data scale. Flores gas generation and the various scenarios. So she tied scenarios pushing through. Scenario based through various scenarios to various scenarios. Or various scenarios have various scenarios with various scenarios. Do scenarios. Various scenarios and various scenarios and the best tech scenarios. Vision. You initiative flow with various scenarios.
; This article is entirely auto-generated using Amazon Bedrock.
Top comments (0)