Welcome to Prof. Saenko's
Computer Vision and Machine Learning Group
at the Computer Science Department of
NEW: Our paper
"Ask, Attend and Answer: Exploring Question-Guided Spatial Attention for Visual Question Answering"
is accepted to ECCV 2016.
NEW: Our paper titled with "Combining Texture and Shape Cues for Object Detection with Minimal Supervision" was accepted to ACCV-16.
NEW: Our paper titled with
"Fine-to-coarse Knowledge Transfer For Low-Res Image Classification"
was accepted to ICIP-16.
Moving to Boston University Kate has recently accepted an Assistant Professor position in the
Computer Science Department at Boston University, and will be moving this summer, along with her group. Stay tuned for the new website.
NEW: Deep CORAL: Correlation Alignment for Deep Domain Adaptation (Extended Abstract).
NEW: Two orals accepted to CVPR 2016:
Deep Compositional Captioning: Describing Novel Object Categories
without Paired Training Data and Natural Language Object Retrieval.
Slides from Kate's MIT talk on March 15th:
Adaptive Deep Learning for
Vision and Language.
Our paper titled Return of Frustratingly Easy Domain Adaptation (Extended Abstract) has won the
of the TASK-CV workshop at ICCV'2015.
Best Paper Prize Our paper titled
Return of Frustratingly Easy Domain Adaptation was accepted to AAAI-16.
Our group is co-organizing the
Transfer and Multi-Task Learning: Trends and New Perspectives Workshop at NIPS 2015 on December 12th, 2015 in Montreal, Canada.
Four papers accepted to ICCV 2015! Here are some spotlights:
Slides from Kate's lecture at the
Microsoft Machine Learning and Intelligence School, which took place in St Petersburg, Russia, are available here:
We will present
Generating Large Scale Datasets from 3D CAD Models at the workshop, initial datasets are available here
I am co-organizing the
Future of Datasets in Vision Workshop at CVPR 2015 on June 11th, 2015 in Boston, MA.
Slides from my recent tutorial on the deep learning library Caffe at the Open Data Science Conference on May 30th in Boston.
Large-Scale Detection by Adaptation 7K Category Detection models are now available!
Our paper titled
From Virtual to Reality: Fast Adaptation of Virtual Object Detectors to Real Domains was accepted to BMVC 2014. See also slides from a recent talk.
DeCAF features that achieve the state of the art on the Office domain adaptation dataset are now available for download.
I am co-chair of the
TASK-CV Workshop on Transferring and Adapting Source Knowledge in Computer Vision, co-located with ECCV2014. Kate Saenko will be giving a
tutorial at CVPR 2012 on Domain Transfer Learning for Vision Applications with Dong Xu and Ivor Tsang. I am co-organizing the
Workshop on Integrating Language and Vision, held at NIPS 2011 in Grenada, Spain.