Péter Vajda
Meta (United States) · 72 works · 4,944 citations · h-index 26
Top works
- FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Search
- Visual Transformers: Token-based Image Representation and Processing for Computer Vision
- Machine Learning at Facebook: Understanding Inference at the Edge
- ChamNet: Towards Efficient Network Design Through Platform-Aware Model Adaptation
- SqueezeSegV3: Spatially-Adaptive Convolution for Efficient Point-Cloud Segmentation
- Open-Vocabulary Semantic Segmentation with Mask-adapted CLIP
- FBNetV2: Differentiable Neural Architecture Search for Spatial and Channel Dimensions
- Value-Aware Quantization for Training and Inference of Neural Networks
- Mixed Precision Quantization of ConvNets via Differentiable Neural Architecture Search
- DSD: Dense-Sparse-Dense Training for Deep Neural Networks
- Cross-Domain Adaptive Teacher for Object Detection
- FBNetV3: Joint Architecture-Recipe Search using Predictor Pretraining
- FBNetV5: Neural Architecture Search for Multiple Tasks in One Run
- Efficient Segmentation: Learning Downsampling Near Semantic Boundaries
- Visual Transformers: Where Do Transformers Really Belong in Vision Models?
- FBNetV3: Joint Architecture-Recipe Search using Neural Acquisition Function
- Towards Fully Automatic Image Segmentation Evaluation
- Cache Me if You Can: Accelerating Diffusion Models through Block Caching
- In Tags We Trust: Trust modeling in social tagging of multimedia content
- One shot 3D photography