Voxel51
Video Understanding: Perceptron AI and FiftyOne
Exploring video-native multimodal AI and integrating cutting-edge video understanding models into computer vision workflows with Akshat Shrivastava.
Open Data Science Conference (ODSC)
Mastering Visual AI with Vision-Language Models & Advanced Evaluation Techniques
Deep dive into inspecting datasets, utilizing VLMs for dataset enrichment, and evaluating fine-tuned models under distribution shifts.
Voxel51
Curating Robotic Data Collection with Voxel51
Addressing the robot learning bottleneck: why data quality and curation matter more than raw volume when collecting embodied episodes.
Voxel51
A Tour of Robotics Datasets: MCAP and LeRobot in Voxel51
Walkthrough of multimodal robotics datasets, authoring MCAP containers, and visualizing state/action spaces inside FiftyOne.
Voxel51
Visualizing and Debugging Robot Learning Data with Voxel51 and Hugging Face LeRobot
Inspecting Hugging Face LeRobot dataset episodes, diagnosing sensor drift, and triaging teleoperation failures.
Voxel51
Flash Demo: Agents in Action with FiftyOne
Demonstrating autonomous AI agents executing computer vision query and triage tasks inside FiftyOne.
Voxel51
Advanced Computer Vision Data Curation and Model Evaluation Workshop
Hands-on workshop covering advanced curation metrics, embedding analysis, and error diagnosis for production vision models.
Voxel51
Finding and Fixing Agricultural Data Annotations with the FiftyOne Agent
Using autonomous agents to detect labeling noise, fix mislabeled bounding boxes, and curate high-quality agricultural CV datasets.
Computer Vision Meetup
Curating Excellence: Strategies for Optimizing Visual AI Datasets
Practical strategies for data-centric AI: dataset pruning, deduplication, and hard negative mining.
Voxel51
Exploring Tokyo’s Most Dangerous Intersection: A Multimodal AV Dataset with Voxel51
Analyzing complex autonomous vehicle edge cases across multiple camera streams and sensor modalities in urban driving environments.
Voxel51
Exploring Video Datasets with FiftyOne and Vision-Language Models
Using VLMs to perform temporal search, natural-language video querying, and automatic captioning.
Voxel51
Workshop: Document Visual AI with FiftyOne—When a Pixel is Worth a Thousand Tokens
Going beyond OCR: parsing complex tables, architectural diagrams, and unstructured layouts with vision foundation models.
Voxel51
Plugins as Products: Bringing Visual AI Research into Real-World Workflows with FiftyOne
Building production-grade UI and backend plugins for computer vision teams to bridge research models into interactive tools.
Voxel51
FiftyOne Computer Vision Plugins: Semantic Document Search
Building custom semantic document search plugins using vector embeddings and multimodal retrieval.
Voxel51
Building Real-World Computer Vision Systems with Voxel51
Architecture patterns for computer vision pipelines: from raw data ingestion to active learning and production monitoring.
Data Phoenix
How to Instruction Tune a Base Language Model
A practical, code-level guide to instruction-tuning base open-weight language models using PEFT and QLoRA.
Data Science Dojo
Understanding and Visualizing ResNets that Forever Revolutionized Deep Learning
First-principles breakdown of residual connections, vanishing gradients, and the architectural mechanics behind deep CNNs.