[HN Gopher] Show HN: Lemon Slice Live - Have a video call with a...
___________________________________________________________________
Show HN: Lemon Slice Live - Have a video call with a transformer
model
Hey HN, this is Lina, Andrew, and Sidney from Lemon Slice. We've
trained a custom diffusion transformer (DiT) model that achieves
video streaming at 25fps and wrapped it into a demo that allows
anyone to turn a photo into a real-time, talking avatar. Here's an
example conversation from co-founder Andrew:
https://www.youtube.com/watch?v=CeYp5xQMFZY. Try it for yourself
at: https://lemonslice.com/live. (Btw, we used to be called
Infinity AI and did a Show HN under that name last year:
https://news.ycombinator.com/item?id=41467704.) Unlike existing
avatar video chat platforms like HeyGen, Tolan, or Apple Memoji
filters, we do not require training custom models, rigging a
character ahead of time, or having a human drive the avatar. Our
tech allows users to create and immediately video-call a custom
character by uploading a single image. The character image can be
any style - from photorealistic to cartoons, paintings, and more.
To achieve this demo, we had to do the following (among other
things! but these were the hardest): 1. Training a fast DiT model.
To make our video generation fast, we had to both design a model
that made the right trade-offs between speed and quality, and use
standard distillation approaches. We first trained a custom video
diffusion transformer (DiT) from scratch that achieves excellent
lip and facial expression sync to audio. To further optimize the
model for speed, we applied teacher-student distillation. The
distilled model achieves 25fps video generation at 256-px
resolution. Purpose-built transformer ASICs will eventually allow
us to stream our video model at 4k resolution. 2. Solving the
infinite video problem. Most video DiT models (Sora, Runway, Kling)
generate 5-second chunks. They can iteratively extend it by another
5sec by feeding the end of the 1st chunk into the start of the 2nd
in an autoregressive manner. Unfortunately the models experience
quality degradation after multiple extensions due to accumulation
of generation errors. We developed a temporal consistency
preservation technique that maintains visual coherence across long
sequences. Our technique significantly reduces artifact
accumulation and allows us to generate indefinitely-long videos.
3. A complex streaming architecture with minimal latency. Enabling
an end-to-end avatar zoom call requires several building blocks,
including voice transcription, LLM inference, and text-to-speech
generation in addition to video generation. We use Deepgram as our
AI voice partner. Modal as the end-to-end compute platform. And
Daily.co and Pipecat to help build a parallel processing pipeline
that orchestrates everything via continuously streaming chunks. Our
system achieves end-to-end latency of 3-6 seconds from user input
to avatar response. Our target is <2 second latency. More
technical details here: https://lemonslice.com/live/technical-
report. Current limitations that we want to solve include: (1)
enabling whole-body and background motions (we're training a next-
gen model for this), (2) reducing delays and improving resolution
(purpose-built ASICs will help), (3) training a model on dyadic
conversations so that avatars learn to listen naturally, and (4)
allowing the character to "see you" and respond to what they see to
create a more natural and engaging conversation. We believe that
generative video will usher in a new media type centered around
interactivity: TV shows, movies, ads, and online courses will stop
and talk to us. Our entertainment will be a mixture of passive and
active experiences depending on what we're in the mood for. Well,
prediction is hard, especially about the future, but that's how we
see it anyway! We'd love for you to try out the demo and let us
know what you think! Post your characters and/or conversation
recordings below.
Author : lcolucci
Score : 183 points
Date : 2025-04-24 17:10 UTC (1 days ago)
| bigyabai wrote:
| > reducing delays and improving resolution (purpose-built ASICs
| will help)
|
| How can you be sure? Investing in an ASIC seems like one of the
| most expensive and complicated solutions.
| lcolucci wrote:
| We wouldn't build it ourselves, but there are several companies
| like Etched, Groq, and Cerebras working on purpose-built
| hardware for transformer models. Here's more:
| https://www.etched.com/announcing-etched
| doublerabbit wrote:
| "Try it now live" and then request me to enter my email.
|
| I'll pass thanks.
| sid-the-kid wrote:
| That's fair. We just removed the sign-in for HN. Should be live
| shortly.
|
| Each person gets a dedicated GPU, so we were worried about
| costs before. But, let' s just go for it.
| yahoozoo wrote:
| Do you use a cloud-based GPU provider?
| sid-the-kid wrote:
| Yes. We use Modal (https://modal.com/), and are big fans of
| them. They are very ergonomic for development, and allow us
| to request GPU instances on demand. Currently, we are
| running our real-time model on A100s.
| lostmsu wrote:
| I see you are paying $2/h. Shoot me an email at victor ta
| borg.games if your model would fit on RTX 3090 24G to get
| it down to $0.2/h (fellow startup).
| tough wrote:
| maybe demos could be a downsampled bitrate/size running
| on commercial GPU's
| ivape wrote:
| How much would this demo cost you from the HN traffic if you
| don't mind me asking?
| sid-the-kid wrote:
| Good question. I guess depends on how many users we get.
| Each users gets their own dedicated GPU. Most video
| generations systems (and AI systems in general) can share
| GPUs during generation. Since we are real time, we don't do
| that. So, each user minute is a GPU minute. This is the
| biggest driver of the cost.
| tough wrote:
| feels like the next logical step for you to bring
| enconomies of scale is to allow users generating the
| video to automatically stream it to n platforms, so each
| gpu can be generating 1 png for many humans to watch
| simultaneously, with maybe 1 human driving the seat on
| what to generate, or more ai, idk
| sid-the-kid wrote:
| that's a good idea! Would be especially cool if the human
| is charismatic and does a good job driving the convo.
| Maybe we can try it out with a streamer.
| tough wrote:
| Vtuber comes to mind
| pjc50 wrote:
| Neuro/vedal already there, although not with the model as
| well.
| sgrove wrote:
| I think it's not going well? I keep getting to the start a
| new call page, it fails, and takes me back to the live page.
| I assume your servers are on fire, but implementing some
| messaging would help ("come back later") or even better, a
| queueing system ("you're N in line") would help a lot.
|
| Really looking forward to trying this out!
| andrew-w wrote:
| We're back online! One of our cache systems ran out of
| memory. Oops. Agree on improved messaging.
| tetris11 wrote:
| If you could lower the email signup for a few hours, that'd be
| nice. I'm not going to sign up for yet another service I'm unsure
| about.
| sid-the-kid wrote:
| We just removed email signup. You can try it out now without
| logging in. It was easier than expected to do technically, so
| we just shipped a quick update.
| tetris11 wrote:
| Thanks! This is amazing
| sid-the-kid wrote:
| Glad you like it! IMO, biggest things to improve on are 1)
| time to video response and 2) being able to generate more
| complicated videos (2 people talking to each other, a
| person walking + talking, scene cuts while talking).
| dang wrote:
| https://lemonslice.com/api/videos/video-XzDwIcW6QCvSIj1vX1Hu...
| lcolucci wrote:
| haha this is amazing! Just made him a featured character. Folks
| can chat with him by searching for "Devil"
| lostmsu wrote:
| This is very impressive. Any details about model architecture and
| size? Input and output representation?
|
| How does voice work? You mentioned Deepgram. Does it mean you do
| Speech-to-Text-to-Speech?
| lcolucci wrote:
| thank you! We have an architecture diagram and some more
| details in the tech report here:
| https://lemonslice.com/live/technical-report
|
| And yes, exactly. In between each character interaction we need
| to do speech-to-text, LLM, text-to-speech, and then our video
| model. All of it happens in a continuously streaming pipeline.
| sid-the-kid wrote:
| For the input, we pass the model: 1) embedded audio and 2) a
| single image (encoded with a causal VAE). The model outputs the
| final RGB video directly.
|
| The key technical unlock was getting the model to generate a
| video faster than real-time. This allows us to stream video
| directly to the user. We do this by recursively generating the
| video, always using the last few frames of the previous output
| to condition the next output. We have some tricks to make sure
| the video stays relatively stable even with recursion.
| tough wrote:
| I'm not at that level but reminded me of
| https://news.ycombinator.com/item?id=43736193
| sid-the-kid wrote:
| Nice find! I hand't seen this before (and will take a
| deeper look later). It looks like this is an approach to
| better utilize the GPU memory. And, we would probably
| benefit from this to get more of a speed-up, which would
| also help us get better video quality.
|
| I do not think they are running in real time though. From
| the website: "Personal RTX 4090 generates at speed 2.5
| seconds/frame (unoptimized) or 1.5 seconds/frame
| (teacache)." That means it would take them 37.5s to
| generate 1 second of video, which is fast for video but way
| slower than real time.
| tough wrote:
| Yep, this is way slower but considered SOTA on video-gen
| open source.
|
| I mostly meant the using the previous frames to generate
| new frames insight that reminded me but lack knowledge on
| the specifics of the work
|
| glad if its useful for your work/research to check out
| the paper
|
| edit: the real-time-ness of it also has to have into
| equation what HW are you running your model on, obviously
| easier to make so on a H100 than a 3090, but these memory
| optimizations really help to make these models usable at
| all for local stuff, which is a great win i think for
| overall adoption/further stuff being build upon them a
| bit like sd-webui from automatic1111 alongside stable
| diffusion weights models being open sourced was a boom on
| image gen a couple years back
| dheera wrote:
| Nice! What infra do you use for inference? I'm wondering what
| the cost-effective platforms are for projects like this. GPUs
| on AWS and Azure are incredibly expensive for personal use.
| sid-the-kid wrote:
| We use modal (https://modal.com/). They give us GPUs on-
| demand, which is critical for us so we are only paying for
| what we are using. Pricing is about $2/hr per GPU (as a
| baseline of the costs). Long story short, things get VERY
| expensive quickly.
| tony_cannistra wrote:
| Nice. This is also how recent advances in ML weather
| forecasting work. Weather forecasting really is just "video
| generation" but in higher dimensions.
| srameshc wrote:
| I am very much fascinated by this virtual avatar talking thing. I
| tried video-retalking https://github.com/OpenTalker/video-
| retalking just to see how far I can make it work to make a
| talking avatar but it is tremendously difficult. But this holds
| tremendous possibilities and I hope it can be eventually cheaper
| to run such models. I know this is far superior and probably a
| lot different but I hope to find open source solutions like Lemon
| Slice someday that I can experiment with.
| sid-the-kid wrote:
| Nice! Thanks for sharing. I hadn't seen that paper before.
| Looks like they take in a real-world video and then re-generate
| the mouth to get to lip synch. In our solution, we take in an
| image and then generate the entire video.
|
| I am sure they will have open source solutions for fully-
| generated real-time video within the next year. We also plan to
| provide an API for our solution at some point.
| sid-the-kid wrote:
| The system just crashed. Sorry! Working on getting things live
| again as fast as we can!
| PUSH_AX wrote:
| Ah the ole HN soak test.
| sid-the-kid wrote:
| Ya. You always think you cross your Ts. But, the law always
| holds.
| lcolucci wrote:
| haha one of the reasons launching on HN is great!
| sid-the-kid wrote:
| We are live again folks! Sorry about that. We ran out of
| storage space.
| gitroom wrote:
| honestly this feels kinda huge - stuff like this is moving so
| fast, it's insane seeing it go real-time
| lcolucci wrote:
| thanks so much for the kind words! we agree that the leap to
| real-time feels huge. so excited to share this with you all
| sid-the-kid wrote:
| IMO, most videos models will be fully real time within 2 years.
| You will be able to pick a model, imagine any world and then be
| fully immersed in it. Walk around any city interacting with
| people, first person shooter games on any map with crazy
| monsters, or just let the model auto-pilot an adventure for
| you.
| genewitch wrote:
| Probably not, but even if so, how much will that cost?
| There's AI that will take a pronpt like "what's the best weed
| whacker in 2025" and build a whole web page to publish the
| review. It's great, awesome. $10 in tokens to do that.
|
| And that's probably still a subsidized cost!
|
| Bfw "what is the best weed whacker" is John C. Dvorak's "AI
| test"
| benob wrote:
| Very nice. Are you planning a paper?
| lcolucci wrote:
| thank you! No concrete paper plan yet as we're focused on
| shipping product features. anything specific you'd want to read
| about?
| elternal_love wrote:
| Hmm, plug this together with a app which collects photos and
| chats with a deceased love one and you have a working Malachim.
| Might be worth a shot.
|
| Impressive technology - impressive demo! Sadly, the conversation
| seems to be a little bit overplayed. Might be worth plugging
| ChatGPT or some better LLM in the logic section.,
| andrew-w wrote:
| Thanks for the feedback. Optimizing for speed meant we had
| fewer LLMs to choose from. OpenAI had surprisingly high
| variance in latency, which made it unusable for this demo. I
| think we could probably do a better job with prompting for some
| of the characters.
| genewitch wrote:
| You know the trick of having Gemini or mistral re-jigger the
| prompt?
|
| Also, you do realize that this will be used to defraud people
| of money and/or property, right?
|
| All about coulda, not shoulda.
| netdevphoenix wrote:
| > Also, you do realize that this will be used to defraud
| people of money and/or property, right?
|
| Sadly, no one cares. LLM driven fraud is already happening
| and it is about to become more profitable.
| aorloff wrote:
| Max Headroom lives !
| sid-the-kid wrote:
| Does he? I can't find him.
| sid-the-kid wrote:
| Looked it up. Cool reference.
| andrew-w wrote:
| Just added as a public character :)
| consumer451 wrote:
| Really wish his trademark glitching head nod was there, but I
| can imagine how that might not be possible.
|
| Super cool product in any case.
| zebomon wrote:
| This is impressive. The video chat works well. It is just a hair
| away from a very comfortable conversation. I'm excited to see
| where you have it a year from now, if it turns out to be
| financially viable. Good luck!
| lcolucci wrote:
| Thank you! Very much agree that we need to improve speed to
| make the conversation more comfortable. Our target is <2sec
| latency (as measured by time to first byte). The other building
| blocks of the stack (like interruption handling, etc) will get
| better in the coming months as well. In 1 year things should
| feel like the equivalent of a zoom conversation with another
| human.
| andrewstuart wrote:
| A really compelling experience.
|
| It seems clumsy to use copyrighted characters in your demos.
|
| Seems to me this will be a standard way to interact with LLMs and
| even companies - like a receptionist/customer
| service/salesperson.
|
| Obviously games could use this.
| ashishact wrote:
| This is just brilliant. Hope you succeed, so that eventually I
| get an API to play with.
| andrew-w wrote:
| Thanks! What kind of use case are you thinking about?
| wouterjanl wrote:
| Really cool stuff. It felt strangely real. Impressive!
| andrew-w wrote:
| Thanks! We think we can cut down the latency to <2s which
| should make it feel even more natural.
| NoScopeNinja wrote:
| Hey, this looks really cool! I'm wondering - what happens if you
| feed it something totally different like a Van Gogh painting or
| anime character? Have you tested any non-photo inputs?
| andrew-w wrote:
| It works with any style of character! Check out the embedded
| videos in our tech report. Peachy and the toilet are my
| favorite. https://lemonslice.com/live/technical-report
| djaychela wrote:
| Just talked with Max Headroom and Michael Scott - my wife is an
| office fan so knows the references, and I know enough Max to ask
| the right things.
|
| Overall, a fun experience. I think that MH was better than Scott.
| Max was missing the glitches and moving background but I'd
| imagine both of those are technically challenging to achieve.
|
| Michael Scott's mouth seemed a bit wrong - I was thinking Michael
| J Fox but my wife then corrected that with Jason Bateman - which
| is much more like it. He knew Office references alright, but
| wasn't quite Steve Carell enough.
|
| The default while it was listening could do with some work, I
| think - that was the least convincing bit; for Max he would have
| just glitched or even been completely still I would think.
| Michael Scott seemed too synthetic at this point.
|
| Don't get me wrong, this was pretty clever and I enjoyed it, just
| trying to say what I found lacking without trying to sound like I
| could do better (which I couldn't!).
| andrew-w wrote:
| Thanks for the feedback. This is definitely a demo where every
| piece matters for maximizing the enjoyment factor. We spent the
| most effort on optimizing video quality and latency, but not a
| lot on tweaking the character prompts that go into the LLM.
| Turns out that matters a lot too.
| bsenftner wrote:
| This is fantastic. I was the founder of the 3D Avatar Store, a
| company that was doing similar things 15 years ago with 3D
| reconstructions of people. Your platform is what I was trying to
| build back then, but at the time nobody thought such tech was
| possible, or they seriously wanted to make porn, and we refused.
| I'll try reaching out through channels to connect with your team.
| I come from a feature film VFX, Academy Award quality work, so it
| would be interesting to discuss. Plus, I've not been idle since
| the 3D Avatar Store, not at all...
| andrew-w wrote:
| We've been very inspired by interactive character experiences
| powered by traditional VFX + puppetry (turtle talk with crush
| is a favorite). I think that sort of interactive entertainment
| will become more commonplace as tech like ours continues to
| improve. Looking forward to connecting!
| inhumantsar wrote:
| love the demo video with Andrew. showing the potential as well as
| the delays and awkwardness of AI is refreshing compared to the
| heavily edited hype reels that are so common
| andrew-w wrote:
| I spent about 2 hours recording videos with different
| characters. Of course, the one I made as a joke for myself and
| never intended to share was the most enjoyable to watch :)
| mentalgear wrote:
| So basically the old open-source live-portrait hooked up with
| audio output. Was very glitchy and low res on my side. btw:
| Wondering if it's legal to use characters you don't have rights
| to. (how do you justify possible IP infringement)
| pjc50 wrote:
| > Wondering if it's legal to use characters you don't have
| rights to. (how do you justify possible IP infringement)
|
| IP law tends to be "richer party wins". There's going to be a
| bunch of huge fights over this, as both individual artists and
| content megacorps are furious about this copyright infringment,
| but OpenAI and friends will get the "we're a hundred-billion-
| dollar company, we can buy our own legislation" treatment.
|
| e.g. https://www.theguardian.com/technology/2024/dec/17/uk-
| propos... a straightforward nationalisation of all UK IP so
| that it can be instantly given away for free to US megacorps.
| andrew-w wrote:
| One way this differs is in the model architecture. Our approach
| relies on a single pass of a diffusion transformer (DiT),
| whereas Live Portrait relies on intermediate representations
| and multiple distinct modules. Getting a DiT to be real-time
| was a big part of our work. Quoting the Live Portrait paper:
| "Diffusion-based portrait animation methods [...] are usually
| [too] computationally expensive." As you hinted at, we had to
| compromise on resolution to get there (this demo is 256x256),
| but we think that will improve over time.
| andrew-w wrote:
| Not relying on facial keypoints means we can animate a wide
| range of non-humanoid characters. My favorite is talking to
| the Doge meme.
| o_____________o wrote:
| Are you going to offer a web embeddable version of the Live
| offering?
| andrew-w wrote:
| It's something we are considering. What use cases do you have
| in mind?
| o_____________o wrote:
| Mostly personal/art projects for now, but willing to pay.
| andrew-w wrote:
| Just added a signup at the bottom of the technical report:
| https://lemonslice.com/live/technical-report
| movedx01 wrote:
| watching baron harkonnen verbally create me code for todo list in
| React was rather amusing, thanks
| andrew-w wrote:
| glad to bring a little joy into the world :)
___________________________________________________________________
(page generated 2025-04-25 23:02 UTC)