Thursday, May 26, 2022
Show HN: I turned my face rec system into a video codec https://ift.tt/08uHQxt
Show HN: I turned my face rec system into a video codec Before the pandemic, my tiny startup was doing quite well selling Edge AI systems, based on our own lightweight AI inference engine, with object detection and face recognition for smart city and smart retail & food service applications. When the real world shut down, there was suddenly nothing to monitor on streets and in restaurants, so I set out to try and evolve our real time face recognition system into a video codec for high quality face-to-face online interactions, as I was not satisfied with the quality of Zoom and friends. I got it to work, and the first release for IOS was just approved on Apple's app store, link: https://ift.tt/jbEU0s5 The way it works is that you create a meeting URL, which you can share out-of-band, for instance via slack or text message. You can also share as a QR code which the app can scan to join a call. You then place your device on a surface in front of you so that the front camera can see you, and it will recognize you face and assign you to your own session, which is broadcast to the meeting channel. If more than one person is in view, both of you will be broadcast but with separate session ids, like if you were on separate cameras. Other meeting participants will show up on your screen and you can start talking. It is optimized for eye contact, meaning that the eyes will actually make it through to the other side as more than just dark pixel clouds, so thinks should feel a bit more personal than the standard Zoom/Teams/or Google Meet call. Because it uses face rec, you can ONLY show your face, and if you disappear from view your audio will stop after a while, to avoid situations like when you need to go the the restroom but forget to mute. This also solves dick-pics etc. The CODEC is not based on H26[45], but is pure AI that runs on the GPU. There is a neural network that compresses the video in real time, and another one decompressing on the receiving end. Finding a tight network architecture that would do this in real time with acceptable quality was a major part of the effort. There are several quality settings possible, but right now it is set fairly high and for 20FPS maxes out around 700kbit/s, though typically uses about half. I've demonstrated good results down to around 200kbit/s, so in theory it should work over satellite links or even Bluetooth. The protocol is UDP with no congestion control but with (Wirehair) FEC to protect against mild packet loss, future versions will detect packet loss and adapt to available bandwidth. The audio just uses OPUS and may click a little bit, I blame AudioEngine or the fact that the last time I wrote audio code was for the game I published for the Amiga in 1994. If you don't have a friend around or multiple devices to play with, there is an "echo test" server mode that allows you to be in a meeting with yourself. Traffic will be peer-to-peer if possible, but otherwise you will be relaying through my tiny Raspberry PI server, so YMMV. I plan to try to switch to something like fly.io soon to improve scalability. There is also a MacOS version coming very soon, and the underlying AI engine also runs on Windows & Linux. Android support is planned. Please take a look and let me know what you think. https://ift.tt/7RDSFxy May 26, 2022 at 02:15PM
Show HN: Arctype, a cross-platform database GUI for developers and teams https://ift.tt/JYoQuhN
Show HN: Arctype, a cross-platform database GUI for developers and teams Hi HN! I’m Justin, founder and CEO of Arctype, and we’re very excited to share Arctype with the community. Arctype is a cross-platform GUI (soon to be open-sourced) to manage and query your databases, with built-in collaboration and visualization. It currently supports Postgres, MySQL, and SQLite, as well as newer-generation databases like ClickHouse, PlanetScale, and Yugabyte (and more coming very soon!). Think of Arctype as “phpMyAdmin meets Postman”. Most of our team grew up learning how to program using the LAMP stack and we missed the experience of interacting with our databases using phpMyAdmin. We wanted to bring back the experience of a simple app that lets you explore and query your databases, but with an updated and modernized interface. At the same time, we were heavily inspired by the ease-of-use and collaboration features of Postman. We wanted to create a super useful, collaborative app like Postman that your whole team uses for development—but for databases instead of APIs. We have a few thousand developers using Arctype today but we haven’t shared it with the HN community yet–we’re excited to hear your feedback! We also have a very active Discord community at arctype.com/discord where developers can ask questions and talk about databases/SQL. I’m sure the community has a lot of questions, so we’ve compiled a list of the most common ones we get: “Why isn’t Arctype open source yet?” It will be soon! We’ve been focusing mostly on features that users have been asking for, as well as performance, stability, and security. We’re not completely happy yet with the documentation and development experience of contributing to Arctype, but this will be a core focus for us in the coming weeks. We just want to make sure it will be very easy for the community to contribute once we publish on GitHub. “How does Arctype make money?” Arctype is free (and will be free forever) for most developers. We have a typical SaaS model for large teams based on a per-seat license as well as an enterprise version that companies can run on their own infrastructure. “Why is Arctype built using Electron?” We wanted to make it easy for anyone to use Arctype–Electron is currently the most practical solution to make it cross-platform and also accessible via a web app. Apps like VSCode and Discord have shown that it’s possible to achieve decent performance so we’re confident that over time the advantages of developing on Electron will outweigh the slight performance hit compared to native apps. “Does Arctype need an account?” You can use Arctype without an account or you can login via email or Google. Certain features such as sharing queries and dashboards with your team require an account. “Is Arctype secure / does it store credentials?” All of your credentials are stored locally, and queries are also executed locally on your machine. We do have a feature that lets you automatically share your credentials to your team, but that is strictly opt-in, and all credentials are encrypted on our backend. Additionally, we have an enterprise version that companies can run on their own infrastructure for organizations that have stringent security requirements. If you’d like to learn more, we have more info at arctype.com/security. “Does Arctype use any analytics software?” Arctype has usage-based analytics and error reporting (we use Sentry) that we use to improve the app and to help us fix bugs/crashes. However, this can be turned off. Additionally, Arctype can be used fully offline. “Why doesn’t Arctype support X database?” Email me at justin at arctype.com :) We’ll make it happen. https://arctype.com/ May 26, 2022 at 12:19AM
Wednesday, May 25, 2022
Show HN: ESP8266 Sprinkler with Runaway Protections https://ift.tt/spC64UZ
Show HN: ESP8266 Sprinkler with Runaway Protections https://ift.tt/rKq9Tzi May 25, 2022 at 07:41AM
Show HN: Run Pi-hole on a local Kubernetes/K3s cluster https://ift.tt/gJ9XDNa
Show HN: Run Pi-hole on a local Kubernetes/K3s cluster https://ift.tt/OcAYQdX May 25, 2022 at 05:21AM
Show HN: A biological circuit simulator built with Godot https://ift.tt/fbkmnT8
Show HN: A biological circuit simulator built with Godot https://ift.tt/fi5FxNT May 25, 2022 at 05:06AM
Show HN: WaifuXL - SOTA image upscaling neural network that runs in the browser https://ift.tt/C9KAkvu
Show HN: WaifuXL - SOTA image upscaling neural network that runs in the browser WaifuXL is our state of the art super resolution model trained on ~1,200,000 anime style images from the Danbooru2021 dataset. You can run it straight from your browser, entirely client-side, at https://waifuxl.com . Using the Onnx Runtime, we stream the weights of our ML models directly onto your device to be executed locally in WebAssembly. Doing so allows us to provide this service solely through a static webpage, no backend for model execution needed. This has the added benefit of enabling the privacy of your images--your images are not, and never will be, sent to us. Our original inspiration for this project was waifu2x ( http://waifu2x.udp.jp ). In general, our model will outperform those used on waifu2x by a significant margin, without the need for multiple models trained on various noise reduction levels. Note that while you _can_ upscale natural (real) images, the model was only trained on anime style drawings, so don’t expect good results. It does work quite well on non-anime graphic art, however. For more information, see our blog post at https://ift.tt/qxRIT1G https://waifuxl.com May 24, 2022 at 11:46PM
Subscribe to:
Posts (Atom)