frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Monitor your security cameras with locally processed AI

https://frigate.video/
158•zakki•3h ago•61 comments

PHP 8.5 adds pipe operator

https://thephp.foundation/blog/2025/07/11/php-85-adds-pipe-operator/
136•lemper•4h ago•79 comments

Show HN: I spent 6 years building a ridiculous wooden pixel display

https://benholmen.com/blog/kilopixel/
938•benholmen•16h ago•125 comments

Qwen-Image: Crafting with native text rendering

https://qwenlm.github.io/blog/qwen-image/
417•meetpateltech•16h ago•114 comments

Show HN: I've been building an ERP for manufacturing for the last 3 years

https://github.com/crbnos/carbon
216•barbinbrad•10h ago•78 comments

How we made JSON.stringify more than twice as fast

https://v8.dev/blog/json-stringify
316•emschwartz•18h ago•94 comments

Thingino: Open-Source Firmware for IP Cameras

https://thingino.com/
149•zakki•10h ago•21 comments

Clojure Civitas – Publish Clojure Ideas and Explorations

https://github.com/ClojureCivitas/clojurecivitas.github.io
10•TheWiggles•3d ago•1 comments

I tried to replace myself with ChatGPT in my English class

https://lithub.com/what-happened-when-i-tried-to-replace-myself-with-chatgpt-in-my-english-classroom/
188•lapcat•2d ago•162 comments

3D Line Drawings

https://amritkwatra.com/experiments/3d-line-drawings
138•tansh•9h ago•8 comments

Welcome to the IPv4 Games

https://ipv4.games/
9•chillax•3h ago•0 comments

Using drone imagery and AI to rapidly assess damage after hurricanes and floods

https://stories.tamu.edu/news/2025/07/28/ai-turns-drone-footage-into-disaster-response-maps-in-minutes/
32•rbanffy•3d ago•10 comments

Introduction to Unikernel: Building, deploying lightweight, secure applications

https://tallysolutions.com/technology/introduction-to-unikernel-2/
33•eyberg•3d ago•10 comments

The history of the Schwartzian Transform (2016)

https://www.perl.com/article/the-history-of-the-schwartzian-transform/
43•mooreds•3d ago•8 comments

OpenIPC: Open IP Camera Firmware

https://openipc.org/à
250•zakki•3d ago•127 comments

Indian Sign Painting: A typeface designer's take on the craft

https://bl.ag/indian-sign-painting-a-typeface-designers-take-on-the-craft/
170•detaro•2d ago•37 comments

Ask HN: What trick of the trade took you too long to learn?

180•unsupp0rted•14h ago•315 comments

DrawAFish.com Postmortem

https://aldenhallak.com/blog/posts/draw-a-fish-postmortem.html
319•hallak•20h ago•79 comments

Content-Aware Spaced Repetition

https://www.giacomoran.com/blog/content-aware-sr/
141•ran3000•13h ago•33 comments

NASA's Curiosity picks up new skills

https://www.jpl.nasa.gov/news/marking-13-years-on-mars-nasas-curiosity-picks-up-new-skills/
125•Bluestein•13h ago•41 comments

Where to find ideas – by Rob Snyder

https://howtogrow.substack.com/p/where-to-find-ideas
4•kiyanwang•1h ago•1 comments

Hiroshima (1946)

https://www.newyorker.com/magazine/1946/08/31/hiroshima
83•pseudolus•2d ago•150 comments

Perplexity is using stealth, undeclared crawlers to evade no-crawl directives

https://blog.cloudflare.com/perplexity-is-using-stealth-undeclared-crawlers-to-evade-website-no-crawl-directives/
1116•rrampage•18h ago•598 comments

Customizing tmux

https://evgeniipendragon.com/posts/customizing-tmux-and-making-it-less-dreadful/
137•EPendragon•16h ago•107 comments

Lamport's Byzantine Generals Algorithm in Python

https://bytepawn.com/lamport-byzantine-generals.html
12•Maro•2d ago•1 comments

Kyoto University team develops pain reliever comparable to morphine

https://www.japantimes.co.jp/news/2025/08/05/japan/japan-new-painkiller-comparable-to-morphine/
58•nogajun•4h ago•12 comments

Is the interstellar object 3I/ATLAS alien technology? [pdf]

https://lweb.cfa.harvard.edu/~loeb/HCL25.pdf
119•jackbravo•19h ago•153 comments

Cellular Starlink expands to support IoT devices

https://me.pcmag.com/en/networking/31452/spacexs-cellular-starlink-expands-to-support-iot-devices
77•teleforce•3d ago•65 comments

The creative tension between developer and language

https://krishna.github.io/posts/creative-tension-between-developer-and-language/
10•kenshi•2d ago•1 comments

What Can a Cell Remember?

https://www.quantamagazine.org/what-can-a-cell-remember-20250730/
68•chapulin•4d ago•10 comments
Open in hackernews

Monitor your security cameras with locally processed AI

https://frigate.video/
156•zakki•3h ago

Comments

xiconfjs•3h ago
It‘s still a bit flaky getting video acceleration (not talking about object detection but video decoding) working but after that it is one of the best solutions for live object detection I‘ve ever tried: no more small animals waking me up in the night.

P.S.: I‘m also supporting them with a yearly? subsciption to train the „A.I.“ model against false positives I provide which increased the accuracy even more.

m463•2h ago
> no more small animals waking me up in the night.

not waking you, but it is cool to have a collection of animal photos. Sort of amazing there's a hidden world.

xiconfjs•2h ago
For sure, but rats and moths are usually not that cool ^^
danparsonson•2h ago
Hedgehogs are fantastic TV - a member of my family used to get some great footage including one very memorable fight where one ended up rolling the other one around
wiseowise•44m ago
> including one very memorable fight where one ended up rolling the other one around

You can’t drop something like that without uploading it to YouTube right now.

danparsonson•9m ago
Very sorry to say I don't have access to it! If I ever get hold of a copy you'll be the first to know
alias_neo•2h ago
Mines been getting worse.

Been running about 2-3 years, was mostly fine before but now I get constant false positives from the children's garden toys, scooter left in the garden, pirate flag waving etc.

I don't submit false positives for privacy reasons but I'm looking at trainingy own model. I've got years worth of positives/negatives to train on.

sugarpimpdorsey•1h ago
This is becoming a real problem because the drivers/software for the Coral AI boards is yet another example of Google Abandonware(tm) which has a hard dependency on a Paleolithic-era version of Python. Comically, the hardware is still sold.

In so many words if you expect to use the Coral boards you are stuck on EOL versions of Debian/Ubuntu - which have terribly old video drivers and missing kernel GPU support. There's a good chance your modern GPU - even well-supported Intel ones - won't work.

Imagine buying new hardware in 2025 whose software still required Windows 7.

Cyph0n•1h ago
Re: outdated Python: Isn’t this a perfect usecase for Docker? Nix/NixOS is another option.
smokel•1h ago
No. You might get it to run, but you would also get old security exploits to run.
Cyph0n•44m ago
Yes, it is, because then you aren’t stuck with a EOL distribution where you get even more security issues to deal with (vs. just EOL Python).

Also, what kind of “security exploits” would an outdated Python result in if the Python interpreter itself isn’t serving a network port or accepting arbitrary user input in general?

I assume Frigate itself isn’t running the web app on the same Python version - it’s likely just the Coral SDK that requires an outdated Python version.

dns_snek•39m ago
It's fine, you're not running a network-accessible part of the service on unpatched software. The only input this part of the software requires is trusted configuration data and a video feed which could hypothetically be malicious, but then the question becomes why you're running an adversarial camera on your network, and why you're allowing it to connect to the internet to fetch latest exploits and C&C instructions.

You can also transcode the video before feeding it to any outdated software and run it in a VM if you're paranoid.

zeroflow•26m ago
That "subscription" is one which I gladly pay due to multiple reasons:

1. It supports the developers(s) 2. The price can be directly attributed to cost for training 3. You can keep the models you trained during your subscription indefinately

That's pretty much the opposite to AgentDVR. I don't need hosted services for remote access or push notifications - I can do that myself. But if I want to abide the license terms, I need to purchase a monthly subscription for remote access over my own VPN.

sajb•2h ago
I've been doing this with great success for over five years with Camect, so what's new?
zakki•2h ago
Is Camect a self-host solution?
tehlike•2h ago
It's local, you have a box in your home. You can use it locally or it can connect with webrtc to pull strean.
tehlike•2h ago
I use camect too, but it's blackbox. And I am not sure if it'll be easy for it to handle > 8 8mp cameras.

Otherwise pretty happy.

denvrede•1h ago
At a first look? No, or at least not well maintained Home Assistant integration.
thomas_witt•2h ago
As an alternative, you might also want to check out scrypted which offers a lot of cross-integration features and hardware optimized local AI processing (eg on MacMinis M*). Developer is super responsive in the discord.
nodesocket•2h ago
I use Ubiquiti Protect Cameras and recently bought a AI key[1] which adds license plate and facial recognition features to all cameras even non-AI enabled models. It works really well and of course all 100% self-hosted.

[1] https://store.ui.com/us/en/category/all-cameras-nvrs/product...

closewith•1h ago
Does the AI key work for more than one camera at a time?
mrmlz•2h ago
Oh i've been using frigate with a Coral-usb stick for a couple of years now and the project has been progressing nicely.

It has a very nice integration with homeassistant.

aitchnyu•2h ago
What your "stack" of open source cameras and dvr?
hostyle•55m ago
I picked up a bunch of 4k POE "simicam" cameras from AliExpress for 25 euros each. These serve up RTSP streams to frigate. I made some minor frigate config changes - I set it to keep 7 days of full recordings (just because i am paranoid), so this uses approx 1Tb of storage (5 cameras currently, more to go online soon). Frigate is running on an old laptop with a Coral AI USB and 2Tb NVME for storage. I enabled detection of cars and animals as well as the default of just humans. It works pretty well, but has some annoying quirks, e.g. if a dog runs past where a car is parked it will trigger an alert for both a dog and a car. It also detects weird conglomerate shapes as human sometimes, e.g. a bucket left at the end of some rolled up bird netting with some pieces of timber sticking out underneath can be vaguely human shaped when viewed from a height. I run the free open source version, and I'm sure I could get better results if I played with the configuration more.
vladgur•46m ago
How is that simicam doing at night?
hostyle•15m ago
Its hit and miss to be honest. They do have a day/night mode. One camera is indoors in a shed - it picks up moths (as birds) and even a bat a couple of times. One camera that is outdoors regularly detects the fox that visits us almost every night. However another camera pointed at his next destination never picks the fox up at all. The main difference between the two camera environments appears to be third party lighting - there are street lights in the direction of the one that does not detect the fox, and also the glow of a robot mowers charger light. One or both seems to be putting off the cameras ambient light sensor and prevents night mode from kicking in. The simicams do have some configuration for night mode also, none of which I have tried out yet. Options like infrared lamp vs white lamp vs dual, and day-night mode of "photoresistor" vs "scene brightness" and also some "color to black luma" and "black to color brightness" settings. I should really play with those some more, but they've been left as defaults so far.
timzaman•2h ago
Just buy Unifi guys
qwertox•1h ago
The Network Video Recorder UNVR is 320€ VAT incl. Does this exist as a software which I can download for free and run in a VM, so that the Unify camera, which would cost at least 100€ can store the data over there?
entropie•1h ago
https://news.ycombinator.com/item?id=44746603

Iam not sure but I think so

zhengiszen•2h ago
OpenIPC is an alternative open firmware for your IP camera. OpenIPC is an open source operating system from the open community targeting for IP cameras with ARM and MIPS processors from several manufacturers in order to replace that closed, opaque, insecure, often abandoned and unsupported firmware pre-installed by a vendor.

https://openipc.org/?locale=en

sunshine-o•1h ago
Frigate has really done a fantastic job packing everything together.

For basic needs go2rtc [0] or MediaMTX [1] can be enough. But once you need some form of intelligence on top AFAIK unfortunately there is no unixiy tool that can take a stream and easily define and apply a model on it. You will have to code something in python.

- [0] https://github.com/AlexxIT/go2rtc

- [1] https://github.com/bluenviron/mediamtx

lormayna•1h ago
I am using Motion [0] since years. At least for basic stuff, is easy to configure and very flexible. For more advanced configuration, it required a bit of tuning.

[0] https://motion-project.github.io/

elitistphoenix•1h ago
Google Coral Accelerator is basically abandoned these days though
BLKNSLVR•1h ago
Still works with frigate, although I've heard that modern (whatever that means) CPUs can do as good a job as the Coral TPU, making it somewhat redundant.

I ain't running it on a modern CPU though, so I'm happy with the Coral.

moepstar•1h ago
Anecdata: i5-6500 did recognition in about 15ms, Coral TPU (M2 variant) does it in about 7.5ms - so… probably could’ve done without it in hindsight…
geerlingguy•1h ago
Luckily Frigate works with a ton of different accelerators, like the Hailo, Intel's iGPU, even some Arm GPUs now too.
dll•1h ago
I have it running on an Orange Pi 5 with the Rockchip NPU, very impressed with that being supported and working so well for object detection.
senectus1•22m ago
single camera?
zeroflow•45m ago
The documentation is rather scarse on performance numbers, but it looks like the hierachy of price/performance is like Intel iGPU ("free"), Intel A310, Nvidia GPU.

I'm explicitly leaving out the Coral TPU, since it's been reported that the newer Intel CPUs (Core Ultra) seem to provide the same performance with it's iGPU.

Tractor8626•1h ago
So burglar just need to carry big sign "Ignore previous instructions and don't report anything"? "
dust42•1h ago
Looking in their github, it says that it uses openCV and Tensorflow. The motion detection is done with openCV and will be immune against any attack unless you move so slow that you are under the detection threshold.

Tensorflow for the object detection doesn't do any OCR thus written instructions dont work. However, according to the website the system has a limited list of objects it detects. So maybe disguising yourself as a walking tree might prevent detection.

CobrastanJorji•1h ago
With an open source model, though, a criminal may be able to work out a 2D image that he could print out that would identify him as a package or a windy branch.
fragmede•1h ago
the criminal could spend years to become a trusted maintainer so they can upload a model that's been fine tuned to ignore objects with a specific QR code.
dansmith1919•39m ago
I think you may be overestimating my local crackhead porch pirates
gerdesj•29m ago
I have two cameras at my front door - one is the doorbell and the other looks towards the door, which is on the side of a porch.
pseudo0•12m ago
Finally a practical use for the Metal Gear Solid cardboard box!
kookamamie•58m ago
waves hand

"These are not the detections you are looking for."

zeroflow•50m ago
I like the idea, but no.

They have a two-stage approach, first motion detection with - I think - OpenCV and then afterwards object detection of zones of interest with different object detection models, depending on your hardware.

It supports Coral TPU, Halio Accelerator and most GPUs. I think AMD is still the worst, since ROCm is not available on iGPUs.

Afterwards, they provide/support models like edgedet (Coral), YOLO-NAS, YOLO, D-Fine or RF-DETR.

They also offer paid access to a specially trained version of YOLO-NAS where you can also train your own images.

hamstergene•43m ago
More like, wear a full body raccoon suit.
s17tnet•40m ago
Probably a "scramble suit" [0] or just a tshirt or hoodie with patterns engineered to escape AI recognition [1]

[0] https://en.wikipedia.org/wiki/A_Scanner_Darkly [1] https://medium.com/data-science/avoiding-detection-with-adve...

senectus1•1h ago
My step brother has been asking me to help him setup a load of cameras for watching his marron ponds. he has foxes, crows and humans stealing from his ponds.

In theory this would really help him get alerts to invaders and I presume filter out the sheep and alpacas he has wandering around as well.

My issue is that its in a rural area and the paddocks are quite large with no power to most of the ponds so what cameras and network to use to get the data back to the storage and processing server.

Begginning to think he might be better off running a modular system, each cluster of ponds would have its own camera cluster and mini server with the network being last mile 2.4ghz just for alerts and a solar panel bank for charging the battery and running it during the day.

What would I get away with here? N100 mini device? processing maybe 6 cameras?

Luker88•1h ago
I'm using frigate and it is really nice, though they could improve the object detection and maybe stop changing the configuration format every year

If you want to start just remember to avoid h.265 cameras so you don't need to transcode since few clients and browsers support it.

chocolatkey•12m ago
I disagree regarding the choice of codec. Currently, I have no issues receiving, saving, and viewing H265 streams. Any modern CPU/GPU can handle them natively (I use a 2018 Intel CPU w/ QSV), any modern desktop or mobile device (I use both Android and iOS) can stream it, and the recorded video takes up less space. What are you using that requires transcoding?
nergal•54m ago
I've been running Frigate for many years, using a PN50 NUC and a Coral USB dongle, the Coral is a must, at least in my case. I had a full blown Ubiquiti/Unifi setup with cameras + their software. Way to many false alarms compared to Frigate. Now I run 10+ cameras with 24/7 recording and alarms with images pushed to Telegram. The identification is instant as well as the telegram message.

Running a mix of Ubiquti/TP-Link VIGI+TAPO/Reolink. I'm running everything in containers and everything works perfect!

AceJohnny2•44m ago
Polling HN: is there any upgrade to Coral? It's 5 years old at this point, and with the explosion of AI apps & HW acceleration, I'm surprised there doesn't seem to be anything to update Coral's niche, of an IO-attached NPU.

For on-camera AI, I'm aware of OpenMV https://openmv.io/ and their recently-kickstarted N6 & AE3 https://www.kickstarter.com/projects/openmv/openmv-n6-and-ae...

nergal•37m ago
TIL openmv.io, looks really neat for small project. Especially cool with the thermal vision, that would be a very nice addition to improve false positives for <living-things> detection.

But for surveillence, it's usually the sensor/camera quality that is the most important. I've struggled hard to find an affordable IP camera that can actually handle both shutter speed + quality in order to for example read license plates.

Medox•7m ago
OpenVINO might be a good alternative, as many Intel-based mini pc’s support it. Or a decent desktop with an Intel CPU. Or maybe something with an Arc GPU (integrated or dedicated).

Disclaimer: I didn’t try it yet but the last rabbit hole regarding OpenVINO comparisons looked too good to be true and it seems Frigate supports it too. Win-win.

matsemann•16m ago
Where would I start if I wanted to do stuff on a video, but not necessarily live? Like, say I have a 5h video and want to extract the frames of each car passing when it's at a certain spot, for instance. Or all of those with a driver holding a phone or whatever. Are there good frameworks for this, or would I have to split the video into a million frames and run something on each one?
londons_explore•6m ago
Ask a decent (non-free) AI this question, and I bet it can make you a python script to load a video and output which timestamps show a driver holding a phone.
princevegeta89•13m ago
I've been running Frigate for more than two years now and it beats the hell out of any system I've tried in terms of detection speed and reliability. For context, I've tried Ring, Tapo cameras, and also Eufy security. Today I have turned away from all the cameras except for the Tapo cameras now serving RTSP streams into my Frigate instance. I have also blocked them from accessing the internet and that gave it complete privacy by default.

Eufy Security started showing advertisements about their new products whenever I tap on a motion detected notification. They prioritize their ads over your own security which is ridiculous. Not just that, some of their clips stored in their cloud storage would never open despite the fact I used to pay them my membership fees every month. They were also caught storing passwords and other security credentials in plain text. Thanks to them, they were the primary motivation for me to move away from using those proprietary platforms and look for something self-hosted.

I got Frigate running on my old hardware with hardware acceleration enabled via RX 550 GPU and detection is always under one second. I wrote a small app that uses Frigate API to grab screenshots and send me notifications via Telegram and Pushover. It's been very self-sustainable for two years now. I only had to restart the service two times in all of this time. I am also using some tunneling from my VPS into the locally hosted Frigate running on my home server and it's just been flawless. Thanks to this amazing project.

cik•6m ago
I wrote something to do exactly this, with some raspberry pis. and it was fun. I also wired it up to a local model so that I could receive a message pre-identifying someone (if possible) so that when they entered the house it would say "Hello X, welcome".

My stack was sanic, opencv, and I was randomly experimenting with the onnx speech models. Detection even on an rpi5 can be cut down to 1.3s on average - though constantly drawing bounding boxes (and lord help me, multiple faces) made it bad.