Michael Haag

Introducing our groundbreaking Labelbox leaderboards: an innovative, scientific process to rank multimodal AI models that goes beyond conventional benchmarks.

Labelbox Leaderboards: Redefining AI Evaluation with Human Experts

Build AI

Labelbox Leaderboards: Redefining AI evaluations with human-centric assessments

Labelbox

Programmatically launch human data jobs for RLHF and evaluation

Rahul Sharma

Evaluating leading text-to-speech models

Metrics-based RAG Development with Labelbox

Raj Jain

Unlocking precision: The "Needle-in-a-Haystack" test for LLM evaluation

Industry: Any

A comprehensive approach to evaluating text-to-video models

Krish Maniar

A comprehensive approach to evaluating text-to-image models

Using Labelbox to improve data quality via AutoQA & advanced labeler review

Using multimodal chat to enhance a customer’s online support experience

Manu Sharma

Working with videos using Gemini 1.5 and multimodal models

Train & fine-tune AI

AI foundations: Understanding embeddings

How to harness AI for efficient video labeling

How to turn LangSmith logs into conversational data with Labelbox

How to use AI to improve website search relevance

How to pre-label images in your annotation project

Use AI

How to accelerate labeling projects using GPT–4V in Foundry

How to improve your task-specific chatbot for better safety, relevancy, and user feedback

How to automate medical imaging with AI

Lisa Dimyadi

Generative AI automates content creation by learning from data. Learn how to implement it and align models with business needs.

How to implement generative AI | Labelbox

How to implement generative AI

Explore & manage data

How to sync your cloud buckets with Labelbox

Building computer vision

How to use AI to automate invoice and document processing

How to accelerate and automate data labeling with labeling functions

Distilling a faster and smaller custom LLM using Google Gemini

Building LLMs

How to enhance RAG chatbot performance by refining a reranking model

How to build equipment detection models to improve worker safety and efficiency

Human-in-the-loop integrates human feedback into AI training, enhancing accuracy and mitigating biases, despite scalability and cost challenges.

What is Human-in-the-Loop? | Labelbox

What is Human-in-the-Loop?

Learn about the process of multimodal data labeling, the technologies used, and the contribution of multimodal data labeling to AI advancements. 

What is multimodal data labeling? | Labelbox

What is multimodal data labeling?

End-to-end workflow with model distillation for computer vision

Knowledge distillation compresses large, powerful AI models into smaller, faster versions without losing performance. Learn how it works.

How to do knowledge distillation | Labelbox

How to do knowledge distillation

How to build generative captioning and enrich product listings faster with foundation models

How to build defect detection models to automate visual quality inspection

How to build defect detection models to improve predictive maintenance

How to build damage classification models with aerial imagery to improve claims automation

Model distillation compresses large models for deployment, transferring knowledge from a teacher to a smaller student model efficiently.

What is Model Distillation?

Learn how AI is integrated to supplement human agents and expedite LLM training using the Reinforcement Learning from AI Feedback (RLAIF) process.

How to Implement Reinforcement Learning from AI Feedback (RLAIF)

RLHF allows users to interactively provide model feedback with corrections, ratings, and preferences. Learn how to implement it with this guide.

How to Implement Reinforcement Learning from Human Feedback

How to Implement Reinforcement Learning from Human Feedback (RLHF)

Learn how you can leverage Labelbox’s Databricks pipeline creator to automatically ingest data from your Databricks domain into Labelbox for data exploration, curation, labeling, and much more.

How to automatically ingest data from Databricks into Labelbox

Learn how to generate human preference data for model comparison or RLHF (reinforcement learning with human feedback) with the new LLM human preference editor. 

How to generate data for model comparison and RLHF

Detecting swimming pools with GPT4 Visual

Learn how to leverage Labelbox's data-centric AI platform to redefine customer care with AI and create solutions tailored to unique customer care challenges.

How to analyze customer reviews and improve customer care with NLP

Learn how to leverage Labelbox’s data-centric AI platform to build a model for content moderation for trust & safety applications. 

How to build a content moderation model to detect disinformation

Industry: Internet & media

Explore how Model Foundry enables teams to efficiently compare and select the right foundation model to kickstart computer vision development, decreasing costs and accelerating time-to-value. 

How to evaluate object detection models with Model Foundry

How to evaluate object detection models with Labelbox Model Foundry

In this guide, we’ll walk through the key difference between these techniques and how to implement them. Explore how to use each of these learning techniques with Labelbox’s LLM Editor & Labelbox Model

Zero-Shot Learning vs. Few-Shot Learning vs. Fine Tuning

Using LLMs

Zero-Shot Learning vs. Few-Shot Learning vs. Fine-Tuning: A technical walkthrough using OpenAI's APIs & models

Introducing the model Foundry - Enrich data and automate tasks using foundation models

How to use the model Foundry for automated data labeling and enrichment

Learn how to harness the power of Model Foundry to automate and enrich data workflows in Labelbox.

Learn how to leverage Labelbox’s data-centric AI platform to build a robust shelf object detection model to boost retail profitability. 

How to boost retail profitability with AI-powered shelf object detection

Industry: Retail & e-commerce

In this guide, we’ll cover how to leverage Open AI’s GPT-3.5 and Labelbox to simplify the fine-tuning process, allowing you to rapidly iterate and refine your models’ performance on specific data.

How to fine-tune OpenAI’s GPT-3.5 Turbo using Labelbox

In this guide, we’ll cover how to leverage Vertex AI and Labelbox to simplify the fine-tuning process, allowing you to rapidly iterate and refine your models’ performance on specific data.

How to fine-tune Vertex AI LLMs with Labelbox

Learn how to leverage Labelbox's data-centric AI platform to build a robust AI-powered product recommendation system to power personalized experiences for retail. 

How to build a powerful product recommendation system for retail

In this guide, we take an in-depth look at the Data I/O process and offer a step-by-step guide to streamline your interaction with the Labelbox platform. 

A guide to the Data I/O process in Labelbox

Explore how to perform model error analysis using Labelbox, with a specific focus on utilizing the Gallery View and Model Metrics.

How to uncover model errors with Gallery View and Model Metrics

How to uncover model errors with Labelbox Model

Learn how to use Meta’s Segment Anything (SAM) model with YOLOv8 to automatically detect, classify, and draw masks on video. 

Using Meta's Segment Anything (SAM) model on video with Labelbox's model-assisted labeling 

In this guide, we'll explore how to use YOLO and Labelbox Catalog together to make a dataset of unlabeled videos searchable.

Leveraging YOLO and Labelbox to make videos queryable by content

Learn how to use Meta’s Segment Anything (SAM) model with YOLOv8 to automatically classify masks. 

Using Meta’s Segment Anything (SAM) model with YOLOv8 to automatically classify masks

Learn how to use BLIP-2-generated captions to create pre-labels for images so a specialized workforce can further improve the image captions.

How to accelerate image-text pair generation with BLIP-2 | Labelbox

Learn how to use BLIP-2-generated captions to create pre-labels for images so that a specialized workforce can further improve the image captions.

How to accelerate image-text pair generation with BLIP-2

Automatically label text with 96%+ accuracy by leveraging Labelbox's search capabilities, bulk classification, and foundation models.

Automatically label text with 96%+ accuracy using foundation models

Learn how to exponentially increase your labeling speed and efficiency by leveraging Labelbox's search capabilities, bulk classification, and foundation models.

Automatically label images with 99% accuracy leveraging Labelbox's search capabilities, bulk classification, and foundation models.

Automatically label images with 99% accuracy using foundation models

Learn how to exponentially increase your labeling speed and efficiency by leveraging Labelbox's search capabilities, bulk classification, and foundation models. 

Leverage an end-to-end system that features the full set of capabilities needed to improve your model’s performance.

Guides | Labeling operations

Labeling automation

Learn how to export your data with more granular control. Filter for specific data rows to export, configure the export to include or exclude information, and more. 

Introducing Export V2: How to export data with more granular control

How to create and label text layers from PDF documents for AI

Learn how to iterate and rapidly fine-tune OpenAI large language models with Labelbox Model. 

How to fine-tune large language models (LLMs) with Labelbox

Learn how to use the Labelbox Connector to quickly explore, organize, and annotate a variety of unstructured data from your Data Lake. 

How to prepare unstructured data for AI and analytics in Databricks

In this guide, we'll show you how you can use foundation models, such as Hugging Face’s embedding extractors, combined with Labelbox’s search capabilities to select impactful data.

Using Labelbox and foundation models to generate custom embeddings and curate impactful data

Learn how you can use model metrics to surface low-performing classes, find and fix labeling errors, and improve the overall performance of the model before it hits production on real-world data. 

An introduction to model metrics

Use Labelbox and Weights & Biases to build better computer vision models with step-by-step workflows of data curation, annotation, & diagnostics

Using Labelbox and Weights & Biases | Labelbox

MLOps

Using Labelbox and Weights & Biases to fine tune your computer vision projects

How to get started in Labelbox Mode: Train evaluate, and improve your ML models

Learn how to ship better models faster by leveraging Labelbox Model. In this guide, we'll walk you through a COCO object detection example to get you onboarded in Model with your first project, model, and model run. 

How to get started in Labelbox Model: train, evaluate, and improve your ML models

How to get started in Labelbox Model: Train, evaluate, and improve your ML models

Learn how to effectively kickstart and scale your data labeling efforts to reduce cost, while maintaining the desired quality required for your use case. 

How to kickstart and scale your data labeling efforts

Learn how to evaluate the results of your labeling project in order to further optimize and improve future iterations and batches of data. 

How to evaluate and optimize your data labeling project's results

Learn how to create a quality strategy to ensure your project is producing valuable, high-quality training data for your use case. 

How to create a quality strategy for your data labeling project

Learn how to effectively define your labeling project's success criteria so all tasks lead to consistent output with high quality. 

How to define your data labeling project's success criteria

A key first step in the labeling process is to align on components of the labeling task. Learn how to define a task, create an ontology, and determine timelines for your labeling project. 

How to define a task for your labeling project

A crucial first step in the labeling process is to align on key components of the labeling task. Learn how to define a task, create an ontology, and determine timelines for your labeling project.

How to define a task for your data labeling project

Find examples of edge cases that will improve model performance. Learn more about how to find similar data in one click with Labelbox. 

How to find similar data in one click | Labelbox

Powerful similarity search capabilities can give your team an edge by helping find specific data points in an ocean of data. Learn more about how to find similar data in one click with Labelbox. 

How to find similar data in one click

Learn how to train a domain-specific chatbot to create custom, tailored experiences for your customers. 

How to train a chatbot | Labelbox

How to train a chatbot

Get started with active learning by leveraging 3 techniques that will help identify what data will most dramatically improve model performance.

Getting started with active learning | Labelbox

Discover how to leverage the 3 active learning techniques that consistently help ML teams more quickly identify what data will most dramatically improve model performance.

Getting started with active learning

Get started with active learning

Learn how you can effectively build an iterative approach to your labeling operations to ensure quality while scaling. 

Scaling up labeling operations while maintaining quality | Labelbox

Many ML teams are eager to label all their data at once. However, this can actually increase time, cost, and quality. Learn how you can effectively build an iterative approach to your labeling operations to ensure quality while scaling. 

How to scale up your labeling operations while maintaining quality

Learn how to maintain quality and cost with Labelbox's Project Performance Dashboard and advanced analytics for Enterprise teams. 

How to maintain quality and cost with advanced analytics

Delays from quality management or the lack of insight into labeling quality can hinder model development. Learn how to maintain quality and cost with Labelbox's Project Performance Dashboard and advanced analytics for Enterprise teams. 

Custom workflows can help optimize how labeled data gets reviewed across multiple tasks and reviewers. Learn more about workflows with Labelbox.

How to customize your annotation review process | Labelbox

Custom workflows can help optimize how labeled data gets reviewed across multiple tasks and reviewers. Workflows is a new feature that allows teams the flexibility to tailor their review workflows for faster iteration cycles. 

How to create a customizable review pipeline to understand the status of your data 

How to customize your annotation review process

Data Rows tab is the central hub for all data rows within a given project. Learn how to better prioritize data to accelerate model development.

How to better manage & search for data with the Data Rows tab | Labelbox

The Data Rows tab is the central hub for all data rows within a given project. You can view, manage, and filter for data rows within your project to better prioritize data for labeling and to accelerate model development.

How to better manage & search for data with the Data Rows tab

How to search, surface, and prioritize data within a project

Queuing data rows with batches gives teams greater control and flexibility in the prioritization of a project’s labeling queue. 

How to prepare and submit a batch for labeling | Labelbox

High-quality training data is crucial to the success of any ML project. Rather than queueing an entire dataset for labeling, queuing data rows with batches gives teams greater control and flexibility in the prioritization of a project’s labeling queue. 

How to queue your data rows with Batches

How to prepare and submit a batch for labeling

A migration guide for the switch to Batch-based queueing, Workflows, and the Data Rows tab.

A new way to queue & review | Labelbox

A new way to queue & review

SDK Changes: A new way to queue & review | Labelbox

SDK Changes: A new way to queue & review

Learn how you can use Labelbox Model to visually compare your ground truths and predictions to identify and fix label errors. 

How to find and fix label errors | Labelbox

How to find and fix label errors

Learn how you can use Model to configure, track, and compare essential model training hyperparameters alongside training data and data splits.

How to curate and version your training datasets and hyperparameters | Labelbox

Learn how you can use Model to configure, track, and compare essential model training hyperparameters alongside training data and data splits. Easily track and reproduce model experiments to observe the differences and share best practices with your team. 

How to curate and version your training datasets and hyperparameters

With the ability to use an NER text layer alongside OCR techniques, teams can annotate text, images, graphs, and more without losing context. 

How to natively annotate a PDF document | Labelbox

In this guide, we'll be walking through how teams can easily turn stores of documents and PDF files into performant ML models with our Document editor. With the ability to use an NER text layer alongside OCR techniques, teams can annotate text, images, graphs, and more without losing context. 

How to natively annotate a PDF document

How to find and fix model errors | Labelbox

A great way to boost model performance is to surface edge cases on which the model might be struggling. You can fix those model failures with targeted improvements to your training data so that the model is better trained on these edge cases. 

Guides

How to set up a delegated access integration between GCP Storage & Labelbox

Introduction to Computer Vision

Data labeling for AI

How to create high-quality image segmentation masks quickly and easily

Best practices for successful image annotation

Try Labelbox today