Awesome breakdown of the challenges of indoor navigation! One thing I was curious about... given that many modern phones now have UWB radios built in, was UWB ever considered as part of the solution stack? From what I understand UWB can get to sub meter accuracy, and I know it's used in several sports applications where precise tracking is really critical.
Is the constraint more about infrastructure (installing anchors, device compatibility, power) or something else that made you lean towards WiFi + SLAM fusion instead?
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.
Fellow BJJ player here – there have been numerous references to the increased risk of vascular injuries and stroke resulting from the high number of chokes that we routinely practice. One study, for instance, found that even brief interruptions in blood flow to the brain during a choke can result in cognitive deficits and other negative health outcomes (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5678212/).
I'm very glad to hear you recovered quickly, and encouraged by the potential for new treatments to help save more lives.
There was no evidence of a vascular dissection in my case. It was either a clot or other piece of debris that made its way to my brain, potentially via a PFO.
I suspect the person you're responding to was just throwing that out carotid dissection as an example; the general risk factor with BJJ is vessel wall injury (which dissection is a subset of). Blood clots from vessel trauma in peripheral parts of the body can migrate to cause thrombosis elsewhere.
Chokes are a big cause of vessel trauma, but hard takedowns, pressure passes, accidental strikes from spazzy white-belts, etc. all count too. Anything that can cause major bruising, really.
To be clear, I'm not trying to argue that anyone should avoid BJJ due to cardiovascular risk, and the sport is probably a large net benefit to cardiovascular health on the whole - but it can certainly be a proximate cause of a stroke if you get unlucky.
Interesting, I hadn't heard about the bit about clots from vessel trauma in peripheral areas.
Thankfully I'm getting my PFO closed in two weeks. The left and right sides of the heart should be isolated systems, and a PFO is a hole that potentially allows a clot to travel from the body side to the brain side. Hopefully if this was related to that, closing the hole will prevent such a thing from happening in the future. Apparently if the hole is closed, a clot would likely travel to the lungs, which are significantly better at coping with it than the brain is.
Does anyone know if there’s a parallel workflow in PyCharm? That is, running on a remote docker container. I haven’t yet been able to get this working but it’d be a vastly superior workflow for my use cases in ML/DS.
There are basically two parts you need to solve. You need files that you edit to save on the remote, and you need to be able to run commands on the remote. So a barebones setup might be SSHFS for mounting files locally, and editing them locally while running commands in an SSH session in your terminal.
Though honestly... you should really try VSCode. You'll get a lot more than simple editing and remote commands (e.g. integrated debugging, etc). VSCode actually installs and runs a headless instance of itself on the remote, and decouples the UI from extensions, language servers, etc. It's a lot more than just editing remote files.
Try downloading it, creating a $5/month VM, and setting it up as a Remote SSH machine. I know it feels like a cult but it's far and away the best editor experience I've had. I switched from Sublime and was up to speed in a day because I could import all my keybindings. You can probably do the same coming from PyCharm.
Hey, don't mean to be promotional, but I just built a platform for email marketing that uses machine learning for automated split testing and optimization (just wrote a post about it here: https://www.optimail.io/blog/move-beyond-ab-testing-email-ca...). I'd be happy to answer any questions about it!
Ooh, neat! I've used GANs to generate synthetic temporal sequence data for training electrophysiologic (i.e., EEG, EMG) signal decoders. In fact, I wrote up the section of my dissertation on this topic today! In my experience it worked quite a bit better than other generative techniques (I've used convolutional variational autoencoders in the past for this and had so-so results). Looking forward to seeing what you guys do with this!
Just finished up my dissertation and I've been working on taking some of the deep learning and RL techniques I've used for my research to build something new.
We just launched Optimail (https://optimail.io) - it uses reinforcement learning to automatically optimize drip email campaigns.
This is really cool, great job. Who is your target market? I've been (slowly) delving into machine learning/AI and found parts of your marketing copy hard to understand, so couldn't imagine somebody with a marketing background being able to grok it.
We only just launched yesterday, so I think we are still working to find that exact product-market fit. At the moment, however, we're targeting medium- to large-sized businesses that use drip email marketing campaigns, such as onboarding campaigns, lead nurturing campaigns, and retention campaigns.
Thanks for the feedback on the copy! If you are feeling extra generous and have a second, shoot me a PM and let me know what parts you found confusing. We want it to be accessible to non-technical marketers who might be using older software like Mailchimp, etc. (After all, a main benefit of using AI here is that you don't need to get into confusing automation-building stuff like multi-branching decision trees.)
Sorry for the late response. I'm a developer and a few sections I had to Google definitions for were:
> Our suite of advanced analytics, including Cohort Analyses, Sequence Analyses, and Algorithm Metrics will unlock new insights in to your customers and campaigns.
> We use a combination of deep reinforcement learning and hierarchical Bayesian models to quickly and accurately learn about your customers and optimize your campaigns in real-time.
I think you could simplify those a bit, maybe hitting from a higher-level with a link to a more detailed explanation of how Optimail does its thing. I totally get the value in using these terms, but for the landing page I would imagine using simpler language may result in non-technical users being able to grok the benefits of your product a bit better.
Thanks for the feedback - it's so hard to hit the right level of description. I'm definitely biased here after just leaving academia since I keep thinking "if we don't mention the hierarchical models they're going to think we're frauds!" - ha.
But you're absolutely right - our goal is to appeal to developers as well as marketing types, so perhaps this level of detail is just confusing and unnecessary on the landing page.
Feel free to shoot me a message if you have any other feedback - I really appreciate it!!
Will do! Btw, I signed up yesterday and really like the look of everything! I'll try and give it a spin in the next month or so whenever I'm ready to set up a campaign.
I'm going to debunk you because as your business partner, I get to constantly annoy you with you loud eating, and this way you have an excuse for being mad.
It's not just the fact that they see a response in AIC that highlights the important aspect of this research, but rather the dissociation between sound evoked activity in controls vs misophonics. While control subjects found unpleasant sounds as annoying as the misophonics found trigger sounds, activity in AIC increased only in the misophonics, and only for the trigger sounds. Interesting to see also that for misophonics, AIC activation scales linearly with the degree of distress - I'm assuming this wasn't the case for unpleasant sounds (but I'm not positive and didn't read closely enough to tell).
You say it's not clear that it's a product of the way their brains are wired, but in the paper they highlight a potential for greater myelination in vmPFC in misophonics. They do this indirectly citing differences in magnetization transfer saturation. I don't know this technique, so I can't really comment on how legit it is, but it seems at least there's some effort to attribute this to structural rather than purely functional differences.
And yeah, of course it's always possible that it's some other factor that is underlying these effects, but I think this dissociation shows reasonably convincing evidence for pursuing more research on this condition (e.g., diffusion imaging to better understand potential structural differences)
Is the constraint more about infrastructure (installing anchors, device compatibility, power) or something else that made you lean towards WiFi + SLAM fusion instead?