News: Research

In deep learning end-to-end training of segmentation is best

Segmentation

A research team (Dr. Guosheng Lin, Prof. Chunhua Shen, Prof. Ian Reid, Prof. Anton van den Hengel) at the School of Computer Science, The 成人大片 developed innovative 鈥淒eep Structured Learning鈥 techniques that set up the new state-of-the-art semantic image segmentation record in the PASCAL VOC Challenge, which is organised by the University of Oxford.  The Adelaide team is the top one currently, outperforming teams from Microsoft Research, Oxford, University of California, Los Angles etc.

[Read more about In deep learning end-to-end training of segmentation is best]

Second Scenes from Video workshop to be held in Chile

SFV

Michael Black and Anton van den Hengel are organising the , which will be held in the Colchagua Valley, one of Chile鈥檚 pre-eminent wine regions, and will include a variety of  functions and excursions exploring the region and its fare.

[Read more about Second Scenes from Video workshop to be held in Chile]

World's best Coco

The Microsoft COCO Captioning Challenge is designed to spur the development of algorithms producing image captions that are informative and accurate.

[Read more about World's best Coco]

AIML semantic image segmentation technique tops the PASCAL VOC Challenge

Researchers at AIML (formerly ACVT) have developed new 鈥淒eep Structured Learning鈥 techniques that set up the new state-of-the-art semantic image segmentation record in the PASCAL VOC Challenge, which is organised by Oxford University. Semantic image segmentation is one of the tasks and probably the most challenging one, which is to label each pixel in images.

[Read more about AIML semantic image segmentation technique tops the PASCAL VOC Challenge]

CVPR 2015 Best Workshop Paper Award

CVPR 2015

In conjunction with the success of the Australian Centre for Robotic Vision鈥檚 workshops at CVPR2015 in Boston:

[Read more about CVPR 2015 Best Workshop Paper Award]

Honourable paper at CVPR 2015

Congratulations to TJ, Pulak, Anders and David. Their paper, Efficient Globally Optimal Consensus Maximisation with Tree Search, recently was awarded a Best Paper Honourable Mention at CVPR.

[Read more about Honourable paper at CVPR 2015]

12 papers at CVPR 2015

The AIML (formerly ACVT) had 12 papers accepted to IEEE Computer Vision and Pattern Recognition 2015 (CVPR鈥15), one of the top 2 conferences in Computer Vision, which must be an Australian record.

[Read more about 12 papers at CVPR 2015]

Best Paper award for Prof Ian Reid at 3DV2014

Professor Ian Reid and colleagues from Oxford have been awarded the Best Paper Award at the International Conference on 3D Vision, held in Tokyo, Japan from 8-11 December 2014

[Read more about Best Paper award for Prof Ian Reid at 3DV2014]

LBT Innovations presents APAS study results

 (ASX:LBT) has published the details of its extensive study into the accuracy of its Automated Plate Assessment System (APAS).

[Read more about LBT Innovations presents APAS study results]

CRC Success

It has just been announced that the Data to Decisions (D2D) CRC application was successful.   The ACVT (now AIML) forms a major node of the D2D CRC, being responsible for the analytics program within the CRC.  The aim of the D2D CRC  is to develop robust data analysis tools applicable to problems of importance to Defence, and other industries.

[Read more about CRC Success]

RSS News Feed