David Vázquez
  • Home
  • Bio
  • Experience
  • Publications
  • Projects
  • Teaching
ESC

Searching...

No results found

↑↓ Navigate ↵ Select
Powered by Hugo Blox
  • Bio
  • Contact
  • Experience
  • Projects
  • Publications
  • Projects
    • AI Tools for Indigenous Languages
    • EnterpriseOps-Gym
    • Apriel Model Family
    • BigDocs
    • WorkArena and BrowserGym
    • SYNTHIA
    • Elektra Autonomous Vehicle
  • News
    • We're hiring researchers and engineers
    • Presenting at ICLR 2026 in Rio de Janeiro
    • NSERC Discovery Grant awarded for Indigenous language AI tools
    • AlignVLM accepted at NeurIPS 2025
    • Appointed Adjunct Professor at Polytechnique Montréal
    • EnterpriseOps-Gym Released
    • StarVector accepted at CVPR 2025
    • BigDocs accepted at ICLR 2025
    • WorkArena accepted at ICML 2024
  • Publications
    • Apriel-1.5-OpenReasoner: RL Post-Training for General-Purpose and Efficient Reasoning
    • StarFlow: Generating Structured Workflow Outputs from Sketch Images
    • VectorGym: A Multitask Benchmark for SVG Code Generation, Sketching, and Editing
    • WildSVG: Towards Reliable SVG Generation Under Real-Word Conditions
    • AgentAda: Skill-Adaptive Data Analytics for Tailored Insight Discovery
    • AlignVLM: Bridging Vision and Language Latent Spaces for Multimodal Document Understanding
    • BigCharts-R1: Enhanced Chart Reasoning with Visual Reinforcement Finetuning
    • BigDocs: An Open Dataset for Training Multimodal Models on Document and Code Tasks
    • Distilling Specialized Orders for Visual Generation
    • Grounding Computer Use Agents on Human Demonstrations
    • Intent Discovery using Large Language Models
    • Rendering-Aware Reinforcement Learning for Vector Graphics Generation
    • StarVector: Generating Scalable Vector Graphics Code from Images and Text
    • UI-Vision: A Desktop-Centric GUI Benchmark for Visual Perception and Interaction
    • WebMMU: A Benchmark for Multimodal Multilingual Website Understanding and Code Generation
    • Improved Training Set Selection for Semi-Supervised Learning
    • A Multimodal Class-Incremental Learning Benchmark for Classification Tasks
    • CADet: Fully Self-Supervised Out-of-Distribution Detection with Contrastive Learning
    • GEO-Bench: Toward Foundation Models for Earth Monitoring
    • Group Robust Classification without Any Group Information
    • InsightBench: Evaluating Business Analytics Agents through Multi-Step Insight Generation
    • RepLiQA: A Question-Answering Dataset for Benchmarking LLMs on Unseen Reference Content
    • Towards Good Validation Metrics for Generative Models in Offline Model-Based Optimisation
    • WorkArena: How Capable Are Web Agents at Solving Common Knowledge Work Tasks?
    • XC-Cache: Cross-Attending to Cached Context for Efficient LLM Inference
    • Anomaly Detection using Graph Neural Networks
    • Automatic Data Augmentation Learning using Bilevel Optimization for Histopathological Images
    • Capture the Flag: Uncovering Data Insights with Large Language Models
    • Constraining Representations Yields Models That Know What They Don't Know
    • Expecting the Unexpected: Towards Broad Out-of-Distribution Detection
    • FigGen: Text to Scientific Figure Generation
    • Flaky Performances When Pretraining on Relational Databases
    • Improving Generalization in Task-Oriented Dialogues with Workflows and Action Plans
    • IntentGPT: Few-Shot Intent Discovery with Large Language Models
    • Knowledge Hypergraph Embedding Meets Relational Algebra
    • Language Decision Transformers with Exponential Tilt for Interactive Text Environments
    • Leveraging Human Preferences to Master Poetry
    • Multilingual Code Retrieval without Paired Data: New Datasets and Benchmarks
    • OC-NMN: Object-Centric Compositional Neural Module Network for Generative Visual Analogical Reasoning
    • OCR-VQGAN: Taming Text-within-Image Generation
    • The Unsolved Challenges of LLMs as Generalist Web Agents: A Case Study
    • TK-KNN: A Balanced Distance-Based Pseudo Labeling Approach for Semi-Supervised Intent Classification
    • 3rd Continual Learning Workshop Challenge on Egocentric Category and Instance Level Object Understanding
    • A Probabilistic Perspective on Reinforcement Learning via Supervised Learning
    • A Survey of Self-Supervised and Few-Shot Object Detection
    • Constraining Low-Level Representations to Define Effective Confidence Scores
    • Contrastive Self-Supervision Defines General-Purpose Similarity Functions
    • CVPR 2020 Continual Learning in Computer Vision Competition: Approaches, Results, Current Challenges and Future Directions
    • Data Augmentation for Intent Classification with Off-the-Shelf Large Language Models
    • Exploring Validation Metrics for Offline Model-Based Optimisation with Diffusion Models
    • Flaky Performances When Pretraining on Relational Databases
    • Implicit Offline Reinforcement Learning via Supervised Learning
    • Multi-Label Iterated Learning for Image Classification with Label Ambiguity
    • OCIM: Object-Centric Compositional Imagination for Visual Abstract Reasoning
    • Overcoming Challenges in Leveraging GANs for Few-Shot Data Augmentation
    • Sequoia: A Software Framework to Unify Continual Learning Research
    • Touch-Based Curiosity for Sparse-Reward Tasks
    • Workflow Discovery from Dialogues in the Low Data Regime
    • 3D Perception with Slanted Stixels on GPU
    • A Deep Learning Localization Method for Measuring Abdominal Muscle Dimensions in Ultrasound Images
    • A Weakly Supervised Consistency-Based Learning Method for COVID-19 Segmentation in CT Images
    • Beyond Trivial Counterfactual Explanations with Diverse Valuable Explanations
    • Decoupling Anomaly Discrimination and Representation Learning: Self-Supervised Learning for Anomaly Detection on Attributed Graph
    • Learning Data Augmentation with Online Bilevel Optimization for Image Classification
    • Seasonal Contrast: Unsupervised Pre-Training from Uncurated Remote Sensing Data
    • SSR: Semi-Supervised Soft Rasterizer for Single-View 2D to 3D Reconstruction
    • Toward Foundation Models for Earth Monitoring: Proposal for a Climate Change Benchmark
    • Weakly Supervised Underwater Fish Segmentation using Affinity LCFCN
    • Counting Objects in Images Based on Approximate Locations
    • A Realistic Fish-Habitat Dataset to Evaluate Algorithms for Underwater Visual Analysis
    • A Weakly Supervised Region-Based Active Learning Method for COVID-19 Segmentation in CT Images
    • Affinity LCFCN: Learning to Segment Fish with Weak Supervision
    • Counting Cows: Tracking Illegal Cattle Ranching from High-Resolution Satellite Imagery
    • Generating Virtual Images for Promoting Visual Artificial Intelligence
    • Instance Segmentation with Point Supervision
    • LOOC: Localize Overlapping Objects with Count Supervision
    • Online Fast Adaptation and Knowledge Accumulation: A New Approach to Continual Learning
    • Online Fast Adaptation and Knowledge Accumulation: A New Approach to Continual Learning
    • Pix2Shape: Towards Unsupervised Learning of 3D Scenes from Images Using a View-Based Representation
    • Proposal-Based Instance Segmentation with Point Supervision
    • Synbols: Probing Learning Algorithms with Synthetic Datasets
    • Adversarial Learning of General Transformations for Data Augmentation
    • Class-Based Styling: Real-Time Localized Style Transfer with Semantic Segmentation
    • Context-Aware Visual Compatibility Prediction
    • Fourier-CPPNs for Image Synthesis
    • Knowledge Hypergraphs: Prediction Beyond Binary Relations
    • Slanted Stixels: A Way to Represent Steep Streets
    • Where Are the Masks: Instance Segmentation with Image-Level Supervision
    • Data for Training Models, Domain Adaptation
    • Environmental Perception for Intelligent Vehicles
    • Learning to Remove Rain in Traffic Surveillance by using Synthetic Data
    • Where Are the Blobs: Counting by Localization with Point Supervision
    • A Benchmark for Endoluminal Scene Segmentation of Colonoscopy Images
    • GPU-Accelerated Real-Time Stixel Computation
    • Guest Editorial: Deep Learning in Computer Vision
    • On-Board Detection of Pedestrian Intentions
    • Semantic Segmentation of Urban Scenes via Domain Adaptation of SYNTHIA
    • Simulation Tools
    • Slanted Stixels: Representing San Francisco's steepest streets
    • The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation
    • Training My Car to See using Virtual Worlds
    • Vision-Based Advanced Driver Assistance Systems
    • Comparison of Two Non-Linear Model-Based Control Strategies for Autonomous Vehicles
    • Embedded Real-Time Stereo Estimation via Semi-Global Matching on the GPU
    • From Virtual to Real World Visual Perception using Domain Adaptation -- The DPM as Example
    • GPU-Based Pedestrian Detection for Autonomous Driving
    • Hierarchical Adaptive Structural SVM for Domain Adaptation
    • Node-Adapt, Path-Adapt and Tree-Adapt: Model-Transfer Domain Adaptation for Random Forest
    • On-Board Object Detection: Multicue, Multimodal, and Multiview Random Forest of Local Experts
    • Pedestrian Detection at Day/Night Time with Visible and FIR Cameras: A Comparison
    • PixelVAE: A Latent Variable Model for Natural Images
    • The SYNTHIA Dataset: A Large Collection of Synthetic Images for Semantic Segmentation of Urban Scenes
    • 3D-Guided Multiscale Sliding Window for Pedestrian Detection
    • Multiview Random Forest of Local Experts Combining RGB and LIDAR Data for Pedestrian Detection
    • Spatiotemporal Stacked Sequential Learning for Pedestrian Detection
    • Vision-Based Offline-Online Perception Paradigm for Autonomous Driving
    • Cost-Sensitive Structured SVM for Multi-Category Domain Adaptation
    • Domain Adaptation of Deformable Part-Based Models
    • Incremental Domain Adaptation of Deformable Part-Based Models
    • Learning a Part-Based Pedestrian Detector in Virtual World
    • Virtual and Real World Adaptation for Pedestrian Detection
    • Adapting a Pedestrian Detector by Boosting LDA Exemplar Classifiers
    • Adapting Pedestrian Detection from Synthetic to Far Infrared Images
    • Computer Vision Trends and Challenges
    • Domain Adaptation of Virtual and Real Worlds for Pedestrian Detection
    • Interactive Training of Human Detectors
    • Learning a Multiview Part-Based Model in Virtual World for Pedestrian Detection
    • Multi-Task Bilinear Classifiers for Visual Domain Adaptation
    • Occlusion Handling via Random Subspace Classifiers for Human Detection
    • Random Forests of Local Experts for Pedestrian Detection
    • Weakly Supervised Automatic Annotation of Pedestrian Bounding Boxes
    • Improving HOG with Image Segmentation: Application to Human Detection
    • Pedestrian Detection: Exploring Virtual Worlds
    • Unsupervised Domain Adaptation of Virtual and Real Worlds for Pedestrian Detection
    • Color Contribution to Part-Based Person Detection in Different Types of Scenarios
    • Cool World: Domain Adaptation of Virtual and Real Worlds for Human Detection using Active Learning
    • Opponent Colors for Human Detection
    • Virtual Worlds and Active Learning for Human Detection
    • Detecting Small Pedestrians
    • Learning Appearance in Virtual Scenarios for Pedestrian Detection
  • Talks & Panels
  • Teaching
  • Team

WorkArena accepted at ICML 2024

2024 · 1 min read

Our paper “WorkArena: How Capable are Web Agents at Solving Common Knowledge Work Tasks?” has been accepted at ICML 2024. Paper · Code

Last updated on Apr 8, 2026
David Vázquez
Authors
David Vázquez
I study how machines learn to act in the world.
← BigDocs accepted at ICLR 2025 Jan 22, 2025

© 2026 David Vázquez. Built with Hugo and HugoBlox.

Made with Hugo Blox. Create your site →