Ethereum

Token

Spot

Ethereum co-founder Vitalik Buterin has endorsed the new Token for Image Tokenizer (TiTok) compression method for its potential blockchain application.
Not to be confused with the social media platform TikTok, the new TiTok compression method significantly reduces image size, making it more practical for storage on the blockchain.
Buterin <a href="https://warpcast.com/vitalik.eth/0x1389d35c" target="_blank" rel="noopener nofollow">highlighted</a> TiTok’s blockchain potential on the decentralized social media platform Farcaster, stating “320 bits is basically a hash. Small enough to go on chain for every user.”
The development could have significant implications for digital image storage of profile pictures (PFPs) and non-fungible tokens (NFTs).
<figure data-ct-non-breakable="true">
 <img src="https://s3.cointelegraph.com/uploads/2024-06/dd298a15-32f1-44b0-aa73-39b82186301d.png" alt="" title="">
 <figcaption style="text-align: center;">
 Source: Thomas
 </figcaption>
</figure>
Related: <a href="https://cointelegraph.com/news/traders-ethereum-most-bullish-altcoin">Traders: Ethereum is the ‘most bullish altcoin’ as ETH reclaims $3.5K</a>
<h2 data-ct-non-breakable="true" id="titok-image-compression">TiTok image compression</h2>
Developed by ByteDance and Technical University Munich researchers, TiTok allows the compression of an image into 32 small data pieces (bits) without losing quality.
According to the TiTok research <a href="https://arxiv.org/pdf/2406.07550" target="_blank" rel="noopener nofollow">paper</a>, advanced artificial intelligence (AI) image compression enables TiTok to compress a 256x256 pixel image into “32 discrete tokens.”
TiTok is a 1-dimensional (1D) image tokenization framework that “breaks grid constraints existing in 2D tokenization methods,” leading to more flexible and compact images.
<blockquote data-ct-non-breakable="true">
 “As a result, it leads to a substantial speed-up on the sampling process (e.g., 410 × faster than DiT-XL/2) while obtaining a competitive generation quality.”
</blockquote>
<figure data-ct-non-breakable="true">
 <img src="https://s3.cointelegraph.com/uploads/2024-06/689a0e6f-5019-45fa-9441-d52e02893ad9.png" alt="" title="">
 <figcaption style="text-align: center;">
 TiTok research paper displaying image compression size comparisons. Source: TiTok
 </figcaption>
</figure>
Related: <a href="https://cointelegraph.com/news/spot-ethereum-etf-ether-eric-balchunas-bloomberg">Spot Ethereum ETFs may begin trading by July 2 — Bloomberg Analyst</a>
<h2 data-ct-non-breakable="true" id="machine-learning-imagery">Machine learning imagery</h2>
TiTok utilizes machine learning and advanced AI, using transformer-based models to convert images into tokenized representations.
The method uses region redundancy, meaning it identifies and uses redundant information in different regions of the image to reduce the overall data size of the end product.
<blockquote data-ct-non-breakable="true">
 “Recent advancements in generative models have highlighted the crucial role of image tokenization in the efficient synthesis of high-resolution images.”
</blockquote>
According to the research paper, TiTok’s "compact latent representation” can yield “substantially more efficient and effective representations than conventional techniques.”
<figure data-ct-non-breakable="true">
 <img src="https://s3.cointelegraph.com/uploads/2024-06/0e92708f-0ab1-4733-8ede-54d0fefe3ed3.png" alt="" title="">
 <figcaption style="text-align: center;">
 Illustration of image reconstruction (a) and generation (b) with the TiTok framework (c). Source: TiTok
 </figcaption>
</figure>
Related: <a href="https://cointelegraph.com/news/vitalik-buterin-nocturne-labs-shuts-down">Vitalik Buterin-backed Nocturne protocol shuts down operations overnight</a>
<h2 data-ct-non-breakable="true" id="titok-not-tiktok">TiTok, not TikTok</h2>
Despite the similar name, TikTok, the social media platform, did not receive endorsement from Buterin.
The Ethereum co-founder, highlighting TiTok’s blockchain potential, lends credibility to the new AI-driven image compression method.
<blockquote data-ct-non-breakable="true">
 “Unlike the existing 2D VQ models that consider the image latent space as a 2D grid, we provide a more compact formulation to tokenize an image into a 1D latent sequence.”
</blockquote>
The proposed new method can “represent an image with 8 to 64 times” fewer tokens than “2D tokenizers,” and the team hopes the research can shed light on “more efficient image representation.”
Magazine: <a href="https://cointelegraph.com/magazine/apple-intelligence-1m-arc-prize-agi-500k-ai-millionaires-ai-eye/">$1M bet ChatGPT won’t lead to AGI, Apple’s intelligent AI use, AI millionaires surge: AI Eye</a>

Vitalik Buterin endorses TiTok AI for onchain image storage

Table of Contents

Cointelegraph

Other articles published on Jun 15, 2024

Best Crypto Presale In June 2024: Investors Shift From Render And Near Protocol To MoonBag

OpenAI reportedly considering shift to for-profit as CEO stacks board

BlockDAG’s Keynote 2 Ignites A Major 1120% Price Spike, Shining With Toncoin And Dogecoin

MoonBag Presale Holds Your Key for Crypto Success, While Cardano and BlastUP Struggle to Keep Up!

Soccer World Cup 2024. Germans defeat Scots 5-1. The ideal beginning

Rebel Satoshi (RECQ) Presale Frenzy Intensifies as Whales Pivot from Shiba Inu (SHIB) and PEPE

Join our free newsletter for daily crypto updates!