正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

building an audio similarity search engine.. given an input track it will find others that sound similar using vector embeddings from a machine learning model with an index of 1.1 million electronic music tracks

hurf

4,141 subscribers

308,625 次观看 • 2 年前 •via X (Twitter)

科学技术音乐

Anya Rossi• Live Now

Private livecam show

10 条评论

hurf 的头像

hurf2 年前

ok u can try it out at still a couple of bugs but hope u enjoy

hurf 的头像

hurf2 年前

to everyone searching for hyph mngo and being disappointed in the results its becuse hyph mngo is unique and nothing else sounds like hyph mngo its the best song ever made

charlotte 🐇 的头像

charlotte 🐇2 年前

this is so sick is there any way i can run a track from my computer through the embedding model and then give ur site the vector and get it to recommend songs to me. also why cosine similarity and not euclidean distance

hurf 的头像

hurf2 年前

thanks! im looking to build a yt link / audio upload -> embeddings feature next which would give recommendations for input tracks not in index. not sure how feasible it would be to do in-browser or locally

frogqueen2008 的头像

frogqueen20082 年前

heartbroken

hurf 的头像

hurf2 年前

sorry ❤️‍🩹

cheese hoover 的头像

cheese hoover2 年前

needs more tracks!! r u gonna try to make more embeddings on ur own with music that wouldn’t have made it into the original set?

hurf 的头像

hurf2 年前

@ramenbois yea! this is a dataset i built from discogs+youtube with more still being ingested also going to add a feature to generate embeddings from a pasted yt link or user upload and return similar tracks for stuff not in index

Joseki 的头像

Joseki2 年前

Any way to turn results into a playlist? It appears the player on the site has it here but Idk how to turn that into an actual playlist.

hurf 的头像

hurf2 年前

good idea! will add a "copy playlist link" button soon since yt playlists can also be made from a list of video ids e.g.

相关视频

New short course on Building Applications with Vector Databases, taught by Pinecone’s Tim Tully! At the heart of a vector database is the ability to store a collection of vectors and then query against that, meaning input a new vector and find similar ones. This is useful for many AI applications. In this course, you'll learn how to use vector databases to build: (i) Semantic Search: Create a text search tool that goes beyond keyword matching, and instead focuses on the meaning of content. (ii) RAG (retrieval augmented generation): Enhance your LLM output by incorporating context from sources the model wasn't trained on. (iii) Recommender System: Combine semantic search and RAG to recommend topics, and demonstrate it with a news article recommender. (iv) Hybrid Search: Build an application that finds items using both images and descriptive text -- by combining both sparse and dense vector representations of the data -- using an eCommerce dataset as an example. (v) Image Similarity: Use image vector embeddings to create an app to compare facial features, using a database of public figures to determine the likeness between them. (vi) Anomaly Detection: Build an anomaly detection app that identifies unusual patterns in network communication logs. I hope you’ll enjoy learning how to build all these types of applications! Please sign up here:

New short course on Building Applications with Vector Databases, taught by Pinecone’s Tim Tully! At the heart of a vector database is the ability to store a collection of vectors and then query against that, meaning input a new vector and find similar ones. This is useful for many AI applications. In this course, you'll learn how to use vector databases to build: (i) Semantic Search: Create a text search tool that goes beyond keyword matching, and instead focuses on the meaning of content. (ii) RAG (retrieval augmented generation): Enhance your LLM output by incorporating context from sources the model wasn't trained on. (iii) Recommender System: Combine semantic search and RAG to recommend topics, and demonstrate it with a news article recommender. (iv) Hybrid Search: Build an application that finds items using both images and descriptive text -- by combining both sparse and dense vector representations of the data -- using an eCommerce dataset as an example. (v) Image Similarity: Use image vector embeddings to create an app to compare facial features, using a database of public figures to determine the likeness between them. (vi) Anomaly Detection: Build an anomaly detection app that identifies unusual patterns in network communication logs. I hope you’ll enjoy learning how to build all these types of applications! Please sign up here:

Andrew Ng

137,073 次观看 • 2 年前

Introducing OpenReasoningEngine, an open-source test-time-compute engine that can be used with any OpenAI-compatible model. Image input, function calling, basic continual learning, + more. This is an early experiment — there are issues that will need to be ironed out. Thread:

Introducing OpenReasoningEngine, an open-source test-time-compute engine that can be used with any OpenAI-compatible model. Image input, function calling, basic continual learning, + more. This is an early experiment — there are issues that will need to be ironed out. Thread:

Matt Shumer

86,967 次观看 • 1 年前

spent my weekend building a niche tool that myself and maybe 100 other music nerds will find useful :) (a simple tool for making rotating vinyl art videos w/ an audio track) - add music - select audio region - add artwork, add background - export video

spent my weekend building a niche tool that myself and maybe 100 other music nerds will find useful :) (a simple tool for making rotating vinyl art videos w/ an audio track) - add music - select audio region - add artwork, add background - export video

jesse

66,602 次观看 • 1 年前

.Weaviate AI Database is an AI-native vector database that allows you to store data objects and vector embeddings from your favorite machine learning models while scaling seamlessly into billions of data objects. AWS and NVIDIA work together in an ecosystem to deliver cost performance and efficiency for Weaviate. Weaviate is available in AWS Marketplace, making deployment within an AWS environment straightforward and accessible.

.Weaviate AI Database is an AI-native vector database that allows you to store data objects and vector embeddings from your favorite machine learning models while scaling seamlessly into billions of data objects. AWS and NVIDIA work together in an ecosystem to deliver cost performance and efficiency for Weaviate. Weaviate is available in AWS Marketplace, making deployment within an AWS environment straightforward and accessible.

AWS Startups

1,472,993 次观看 • 16 天前

Building an Agentic Search System Building an agentic system is not too hard. Loops, function calling, tool execution, and the model. That's it! I show in this video how to build a search agent from scratch. ~350 lines of code!

Building an Agentic Search System Building an agentic system is not too hard. Loops, function calling, tool execution, and the model. That's it! I show in this video how to build a search agent from scratch. ~350 lines of code!

elvis

57,290 次观看 • 1 年前

Greetings Now (My) the Pythia Search Engine™️ v000.3 has a an interface that I've been using locally to test the quality of the results Know that your favorite search engine is probably not searching and uses third-party APIs This is a local test with ~2000 our of many million

Greetings Now (My) the Pythia Search Engine™️ v000.3 has a an interface that I've been using locally to test the quality of the results Know that your favorite search engine is probably not searching and uses third-party APIs This is a local test with ~2000 our of many million

wavefnx

101,900 次观看 • 1 个月前

MusicHiFi Fast High-Fidelity Stereo Vocoding Diffusion-based audio and music generation models commonly generate music by constructing an image representation of audio (e.g., a mel-spectrogram) and then converting it to audio using a phase reconstruction model or vocoder.

MusicHiFi Fast High-Fidelity Stereo Vocoding Diffusion-based audio and music generation models commonly generate music by constructing an image representation of audio (e.g., a mel-spectrogram) and then converting it to audio using a phase reconstruction model or vocoder.

AK

27,285 次观看 • 2 年前

Bob Moog explaining the basics of analog synthesis, from a 1980s BBC documentary. The architect of the Moog synthesizer, an instrument that fundamentally altered the sound of popular music, from jazz to funk, soul, and electronic music as we know it today.

Bob Moog explaining the basics of analog synthesis, from a 1980s BBC documentary. The architect of the Moog synthesizer, an instrument that fundamentally altered the sound of popular music, from jazz to funk, soul, and electronic music as we know it today.

Linear Labs

10,272 次观看 • 1 个月前

New short course: Building Multimodal Search and RAG", by Weaviate AI Database's Sebastia(N_) Witalec ✊🏽✊🏾✊🏿. Contrastive learning is used to train models to map vectors into an embedding space by pulling similar concepts closer together and pushing dissimilar concepts away from each other. This technique is also used to train multimodal embedding models that capture semantic similarity across different modalities like text, images, and audio. These multimodal embeddings can be used to build multimodal search and RAG systems. In this course, you'll learn how contrastive learning works, and how to add multimodality to RAG – so your models can draw on diverse, relevant context to answer questions. For example, a query about a financial report might synthesize information from text snippets, graphs, tables, and slides. You will also learn how visual instruction tuning lets you integrate image understanding into language models, and build a multi-vector recommender system using Weaviate’s open-source vector database. Please sign up here:

New short course: Building Multimodal Search and RAG", by Weaviate AI Database's Sebastia(N_) Witalec ✊🏽✊🏾✊🏿. Contrastive learning is used to train models to map vectors into an embedding space by pulling similar concepts closer together and pushing dissimilar concepts away from each other. This technique is also used to train multimodal embedding models that capture semantic similarity across different modalities like text, images, and audio. These multimodal embeddings can be used to build multimodal search and RAG systems. In this course, you'll learn how contrastive learning works, and how to add multimodality to RAG – so your models can draw on diverse, relevant context to answer questions. For example, a query about a financial report might synthesize information from text snippets, graphs, tables, and slides. You will also learn how visual instruction tuning lets you integrate image understanding into language models, and build a multi-vector recommender system using Weaviate’s open-source vector database. Please sign up here:

Andrew Ng

104,371 次观看 • 2 年前

#CHANYEOL : "upside down" actually means "i'm in a mess right now". it contains lyrics about how my life feels like a mess but I want to turn it around. the 6 tracks range from a wide variety of genres. the title track is an electronic rock track with a bit of rock energy and a powerful guitar sound. i've been working hard preparing for it

#CHANYEOL : "upside down" actually means "i'm in a mess right now". it contains lyrics about how my life feels like a mess but I want to turn it around. the 6 tracks range from a wide variety of genres. the title track is an electronic rock track with a bit of rock energy and a powerful guitar sound. i've been working hard preparing for it

ꓶNꓶN

18,626 次观看 • 11 个月前

Today we’re launching Stable Audio 2.5: The first audio model built for enterprise-grade sound production 🔊 Audio influences brand engagement by 86%, but few enterprises are leveraging audio as an extension of their brand, making customized sound an untapped differentiator. Stable Audio 2.5 is purpose-built for this opportunity to create customizable, high-quality audio at scale, with capabilities that include: ▶️ Improved musical composition: Generate full songs with multi-part structure, meaning a clear intro, middle, and outro. ▶️ Audio inpainting: Input audio, select where the track should start, and the model uses the context to generate the rest of the track. ▶️ Customization: Our team can fine-tune Stable Audio 2.5 to help enterprises create the right sound for their brand. ▶️ Faster inference: The model can generate up to three-minute long tracks in under two seconds on a GPU, outputting in just eight steps (compared to ~50 in the previous model). You can learn more here 👉

Stability AI

67,552 次观看 • 10 个月前

TURN CLAUDE CODE INTO AN AI SEO MACHINE. From content and keyword workflows to full search optimization, this is the kind of setup that can turn Claude into a real traffic engine.

TURN CLAUDE CODE INTO AN AI SEO MACHINE. From content and keyword workflows to full search optimization, this is the kind of setup that can turn Claude into a real traffic engine.

0xMarioNawfal

63,961 次观看 • 3 个月前

We've raised $6.5M to kill vector databases. Every system today retrieves context the same way: vector search that stores everything as flat embeddings and returns whatever "feels" closest. Similar, sure. Relevant? Almost never. Embeddings can’t tell a Q3 renewal clause from a Q1 termination notice if the language is close enough. A friend of mine asked his AI about a contract last week, and it returned a detailed, perfectly crafted answer pulled from a completely different client’s file. Once you’re dealing with 10M+ documents, these mix-ups happen all the time. VectorDB accuracy goes to shit. We built HydraDB for exactly this. HydraDB builds an ontology-first context graph over your data, maps relationships between entities, understands the 'why' behind documents, and tracks how information evolves over time. So when you ask about 'Apple,' it knows you mean the company you're serving as a customer. Not the fruit. Even when a vector DB's similarity score says 0.94. More below ⬇️

We've raised $6.5M to kill vector databases. Every system today retrieves context the same way: vector search that stores everything as flat embeddings and returns whatever "feels" closest. Similar, sure. Relevant? Almost never. Embeddings can’t tell a Q3 renewal clause from a Q1 termination notice if the language is close enough. A friend of mine asked his AI about a contract last week, and it returned a detailed, perfectly crafted answer pulled from a completely different client’s file. Once you’re dealing with 10M+ documents, these mix-ups happen all the time. VectorDB accuracy goes to shit. We built HydraDB for exactly this. HydraDB builds an ontology-first context graph over your data, maps relationships between entities, understands the 'why' behind documents, and tracks how information evolves over time. So when you ask about 'Apple,' it knows you mean the company you're serving as a customer. Not the fruit. Even when a vector DB's similarity score says 0.94. More below ⬇️

Nishkarsh

3,862,274 次观看 • 4 个月前

Coded with Grok 4 from xAI , built this Three.js demo using a 3D model from its examples. Imagine building on it for fun school learning or an immersive game/web experience. Grok 4's coding abilities open up creative possibilities for everyone.

Coded with Grok 4 from xAI , built this Three.js demo using a 3D model from its examples. Imagine building on it for fun school learning or an immersive game/web experience. Grok 4's coding abilities open up creative possibilities for everyone.

Techartist

58,303 次观看 • 1 年前

I can search a database with 32 million records in less than 50 milliseconds. This speed is insane! There's no reason your search is slow. There's no reason you aren't using this library when building applications with a lot of data. Take a look at typesense |. It's an open-source search engine. It's crazy fast and easy to use. They are sponsoring this post. Here is their GitHub repository: You can use the library in Python, JavaScript, PHP, Ruby, Swift, and pretty much any other programming language out there. Here are some of the things you can do with Typesense: • Vector search over embeddings • Conversational search for RAG applications • Image similarity search • Voice search And it does all of this extremely fast. A few interesting notes about the library: • Optimized for developer happiness • Works out of the box for most use cases • Very intuitive API • Easy to deploy and scale Probably the best way to compare them is against ElasticSearch. The latter is much more complex to configure and run with a steep learning curve. Check their GitHub repository. They have examples, tutorials, and everything you need.

I can search a database with 32 million records in less than 50 milliseconds. This speed is insane! There's no reason your search is slow. There's no reason you aren't using this library when building applications with a lot of data. Take a look at typesense |. It's an open-source search engine. It's crazy fast and easy to use. They are sponsoring this post. Here is their GitHub repository: You can use the library in Python, JavaScript, PHP, Ruby, Swift, and pretty much any other programming language out there. Here are some of the things you can do with Typesense: • Vector search over embeddings • Conversational search for RAG applications • Image similarity search • Voice search And it does all of this extremely fast. A few interesting notes about the library: • Optimized for developer happiness • Works out of the box for most use cases • Very intuitive API • Easy to deploy and scale Probably the best way to compare them is against ElasticSearch. The latter is much more complex to configure and run with a steep learning curve. Check their GitHub repository. They have examples, tutorials, and everything you need.

Santiago

160,154 次观看 • 1 年前

This is a big day. Meta is open-sourcing AudioCraft. You can now generate incredible music and sounds with a single prompt. It includes the most performant Generative AI Model (audio) on the market, the "Llama" of Audio. The research framework contains the weights and code of these models: ▸ MusicGen: controllable text-to-music model. ▸ AudioGen: text-to-sound model. ▸ EnCodec: high fidelity neural audio codec. ▸ Multi Band Diffusion: An EnCodec compatible decoder using diffusion. This is going to tremendously speed up audio research 👏

This is a big day. Meta is open-sourcing AudioCraft. You can now generate incredible music and sounds with a single prompt. It includes the most performant Generative AI Model (audio) on the market, the "Llama" of Audio. The research framework contains the weights and code of these models: ▸ MusicGen: controllable text-to-music model. ▸ AudioGen: text-to-sound model. ▸ EnCodec: high fidelity neural audio codec. ▸ Multi Band Diffusion: An EnCodec compatible decoder using diffusion. This is going to tremendously speed up audio research 👏

Lior Alexander

231,608 次观看 • 2 年前

🔉 Introducing SAM Audio, the first unified model that isolates any sound from complex audio mixtures using text, visual, or span prompts. We’re sharing SAM Audio with the community, along with a perception encoder model, benchmarks and research papers, to empower others to explore new forms of expression and build applications that were previously out of reach. 🔗 Learn more:

🔉 Introducing SAM Audio, the first unified model that isolates any sound from complex audio mixtures using text, visual, or span prompts. We’re sharing SAM Audio with the community, along with a perception encoder model, benchmarks and research papers, to empower others to explore new forms of expression and build applications that were previously out of reach. 🔗 Learn more:

AI at Meta

1,250,395 次观看 • 7 个月前