AndrewHart

The only result I get in Google for this is... this post

This is something I built while learning RL, and I decided to open-source. I noticed that every major RL algorithm (Actor Critic, A2C, PPO, TD3 etc.) would be written from scratch, even though they shared a lot of the same features. And every implementation of the same features would be slightly different across each implementation. So trying to take a feature from one algo to another, or even to try building your own features, was a massive pain and error-prone.

HelloRL is a new modular framework, built around a single `train()` function, which scales up to every algorithm. The difference between Actor Critic (discrete, online, monte-carlo, simple) and TD3 (continuous, offline, 1-step rollout, targets, reference critic, etc.) is just a different set of modules. Easy to swap between algorithms, or mix and match features, or build your own modules.

It's great, astounding, divine, amazing, splendiferous.

There are plenty of demo videos on Twitter:

https://x.com/colin_d_m

The motion tracking system used by ARKit/ARCore is called SLAM, combination of sensors + camera data if available.

Great question. Partly on device compatibility, but mostly that we can achieve an amazing, reliable experience right now without those anchors. If we weren't getting the accuracy we needed from our current methods, we would have to look at alternatives like UWB and consider the trade-offs.

We've had the solution live in stores for years without any issue. Our algorithms can clean up and filter out unreliable data, based on all of the other data we have, from other APs, from previous data etc..

If they re-fit the WiFi, which they might spend a month doing once every 5-10 years, it would need a re-survey, which could take 1-2 hours. But thankfully no new infrastructure.

Thank you! I tried to keep it interesting but not get lost in rabbit holes.

We have around 300 hours of ground-truth data now, in 1-second intervals, which we use for algorithm training and refinement. The same as performing a survey, our team marks their location on the map, then walks to the next location, and through post-processing we can correct any errors and interpolate the locations in between. You can see this process in the diagram with the large black dots, where the user marked their location.

Thanks for the feedback. I wanted to keep it balanced to be accessible but also insightful.

To answer your point: we have the digital map, can use that to understand obstacles etc in the space. In some of those larger stores you see in the visual, we typically survey the entire store within 2-3 hours, it's low-effort work, not a blocker.

All of the videos shown on the speed-run of our technology is on 4 year-old Android devices, such as Samsung S21, Pixel 6 etc.. We always test and gather statistics on older devices to fairly represent what's available, rather than latest-and-greatest.

Thanks. Very interested in robotics so that'd be an ideal home, given the SLAM + sensors tech stack.

Yes - one of the limitations with mobile is engaging the camera/quality of the SLAM. With a robot, they're already using SLAM with strong tracking, and controlling the hardware stack means no device limitations (uncalibrated sensors, limited WiFi pings etc.)

Aside from saving time, I'm bad at writing. Especially emails. I often open ChatGPT, paste in the whole email chain, write out the bullets of the points I want to make and ask it to draft a response which frames it well.

You could say the same about software and app stores. If safety were the top priority, then the safest option is to say no apps, but that isn't competitive or lucrative. Apple's approach is to create safe frameworks and a review process that allows the App Store to exist.

Do you track how many people clicked each confetti button?

Fun at parties.

Dent Reality | Android Developer | Remote, Europe (office in London) | https://careers.DentReality.com

Dent Reality is building the digital layer for the physical world. Checkout this 1 min AR demo:

https://twitter.com/AndrewHartAR/status/1543987624016318465?...

We're hiring Android developers to help us build a native SDK in Kotlin. More info here:

https://careers.dentreality.com/jobs/1851300-senior-android-...

Dent Reality | London | Mostly remote, some onsite | https://careers.DentReality.com

Dent Reality is building the digital layer for the physical world. Checkout this 1 min AR demo:

https://twitter.com/AndrewHartAR/status/1499061786565386243

We're hiring across all platforms, and ops:

- Head of Operations

- iOS devs

- Android devs

- Web devs

Monzo is another company which does this incredibly well, and wrote a public document on it: https://monzo.com/tone-of-voice/

I'd love to read more on this topic - how companies learn to do this, and then go about establishing it internally, so that it becomes part of their brand and culture.

Looks promising. One feature I'd need is ability to change the resolution to vertical - the same as the video I'm inputting.

I run an AR startup called Dent Reality - Apple lists us in their Indoor Maps program docs. We've built on IMDF to solve this for grocery stores, and provide shopping list navigation with maps and AR. 60 second demo video here:

https://twitter.com/AndrewProjDent/status/131171456565527347...

Dent Reality | JavaScript Developer | London, UK | Full-Time https://DentReality.com

We’re a “smart spaces” company, building an AR platform that connects people with retail stores and shopping centres, to provide an Amazon-level customer experience.

Our web platform enables our customers to build their own indoor maps, view hotspot analytics, and manage their locations. We’ve previously built the most popular mobile AR project on GitHub (4k stars), and we’re actively working with major retailers around the world to integrate with their locations.

The role is full-time, working from our office in Old Street, London, with the option to spend a couple of days at home each week.

Apply on AngelList or email hiring@dentreality.com https://angel.co/company/dentreality/jobs/653173-javascript-...

I swear by Source Code Pro.

Dent Reality | Full-Stack Developer, Senior iOS Developer | London, UK | Full-Time, ONSITE https://DentReality.com

About us: Dent Reality is building an AR platform to connect people with retail stores and other public spaces, giving them an Amazon-level customer experience. We’re an early-stage startup that’s already developed the most popular open-source project for mobile AR (4k+ stars on GitHub), enabling any developer to build world-scale AR apps:

https://github.com/projectdent/arkit-corelocation

Roles:

Full-Stack Developer: Build self-serve tools, so our customers can set up their spaces and manage their data. You’ll have the opportunity to lead the web team as it grows.

https://angel.co/company/dentreality/jobs/622521-full-stack-...

Senior iOS Developer: Lead development of our mobile app, adding new features and capabilities. Should be experienced in architecture and UI implementation. ARKit experience not required, but a good bonus.

https://angel.co/company/dentreality/jobs/636225-senior-ios-...

Both roles offer salary + equity. You’ll be given ownership and freedom over development, and will be responsible for making key technology, infrastructure, and product decisions.

"researchers found 24 popular iPhone apps that were collecting location data — like Bluetooth beacons to Wi-Fi network names... and cell network names."

A point of note - finding the names of wireless routers, or cell network - requires calling private APIs, so those apps should be banned from the store on that basis alone.

I tried every note-taking and distraction-free writing tool there was - Evernote, iA Writer, ByWord etc.. never stuck with one for more than a few months.

Then I found Ulysses. I love the indexing of all the notes within folders, and the simplicity of the app.

I'm not a huge fan of the painfully slow pace of updates, especially when they're charging a subscription fee. The app is great though.

The problem with Notion is that on iOS it's a web-app, so it feels terrible to use. It white-screens while it reloads whenever you do any navigation, and doesn't support multi-tasking on iPad.

Fonts are expensive. This would be awesome if not an April Fools joke.

I made my own virtual pet game for Apple Watch (where I think they belong, in modern times), a few months ago.

https://projectdent.com/raise-emoji-81a10a67b826

I made a modern virtual pet, for iOS and Apple Watch. Apple Watch seems like the perfect modern form factor for one of these.

https://itunes.apple.com/us/app/raise-emoji/id1045353760

Thanks for the suggestions. I'm taking those on board for updates. And thanks for trying it out.

For the News channel - I tried to get the right mix of being enough, but not too much. Maybe even allowing the user to set the density might be a good idea!