Post has attachment
Amazing video footage of the human #retina, as taken with #epiCam video ophthalmoscope.

Post has attachment
What Does USB 3.1 have to Offer for Industrial Imaging?

Read more about the details of the emerging standard that is currently rolled out in consumer IT and electronics and its impact on the Machine Vision industry. Please share your view by contributing comments at the end of the post!

Post has attachment
BMVA Technical Meeting: Computer Vision in Remote Sensing
One Day BMVA symposium in London, UK 
Wednesday 23rd March 2016 
Chairs: Hong Wei and James Ferryman, University of Reading
Keynote speakers: 1.   Uwe Stilla (Chair of ISPRS Inter Commission Working Group III/VII: Pattern Analysis in Remote Sensing, Technische Universitaet Muenchen); 
Call for contributions:
The workshop aims to bring together researchers in the combined field of computer vision and remote sensing to share ideas, outputs, and challenges for future research. It will feature a number of invited speakers, and a number of talks by interested parties. The work can be “in progress” or recently published, in addition to novel research. If you would like to give a talk, please contact Hong Wei ( by 13th January 2016 with an abstract of your work.
Exploitation of spatial data to support earth observation is significant interest to the computer vision research community. With increasing amounts of spatial data becoming available, and with rapid development of computational technology and remote sensing technology, there are demands to explore new algorithms which can automatically transform raw data to semantically meaningful information. Image analysis and computer vision technology will play an important role in this transformation.
In this workshop we hope to seek the opinions of speakers and participants from a wide range of interesting academics and companies with an interest in computer vision for remotely sensed data analysis and applications. What demands are there from the end-user point of view to make use of spatial data? What academic challenges remain? What elements are ready for adoption by industry? What are the key factors blocking progress? What new application opportunities can be explored?
Areas of interest covered by this workshop include (but not limited to):
•          Airborne and spaceborne remotely sensed data analysis
•          Scene understanding and classification from remotely sensed data
•          Fusion techniques in multi-spectral sensors
•          Change detection from remote sensing
•          Automated feature extraction for segmentation and object recognition in remote sensing
•          Data-driven scene modelling for remote sensing
•          3D scene reconstruction from remotely sensed data
•          Camera calibration, image registration in remote sensing
Book online at
• £10 for BMVA members 
• £30 for non-members 
including lunch

Post has attachment
Good morning and welcome back to BMVA social feeds for 2016. It's time to start the new year by catching up with the latest technology at CES in Las Vegas. Look out for the computer vision apps...

Post has attachment
BMVA Technical Meeting: Deep Learning for Computer Vision 
One Day BMVA symposium in London, UK 
Friday 8th July, 2016 
Chairs: Kai Arulkumaran, Anil Bharath 

Keynote speakers: Andrea Vedaldi (University of Oxford), Ben Graham (University of Warwick), Samim Winiger (, Sander Dieleman (Google DeepMind), Soumith Chintala (Facebook AI Research) 

Call for contributions 
The programme will feature a range of keynote speakers. In addition there will be poster sessions for interested parties. If you would like to bring a poster, please submit the proposed title and abstract via by Friday 15th April, 2016. 

Deep learning is one of the hottest topics in computer science research right now, with numerous publications coming from both academia and industry. Since the success of Alex Krizhevsky’s Convolutional Neural Network (CNN) on the ImageNet classification benchmark in 2012, deep neural networks have begun to dominate different areas in computer vision. Nowadays, some consider object classification to be an almost-solved problem, or at least in the regime of large amounts of data. Similarly, image segmentation has been massively improved through deep learning techniques. However, the lack of labelled data for supervised learning prohibits the application of large, vanilla CNNs to certain problem domains. Recent work has focused on how to tackle these more difficult challenges by combining traditional computer vision techniques with deep learning, as well as advancing more traditional deep learning methods themselves. The aim of the workshop is to explore not just the areas of computer vision that deep learning can advance, but also how the investigation of different areas in computer vision has led to advances in deep learning. On the one hand, more sophisticated architectures and training methodologies can be used to tackle more complex problems in computer vision. On the other hand, the domain of spatial and temporal dimensions in computer vision can be more intuitively explored by humans, leading to a better understanding of what, and possibly why, deep neural networks learn what they do. In particular, visualisation techniques have great potential in helping us understand both the models and the data that we work with. 

Topics of Interest 
With respect to computer vision, the areas of interest covered by this meeting include the following: 
• Neural network architectures 
• Deep learning optimisation techniques 
• Visualisation of learned representations 
• Practical applications of deep learning 
• Generative neural network models 
• Hybrid traditional CV/deep learning systems 

Book online at
• £10 for BMVA members 
• £30 for non-members 
including lunch

#MachineVision #Technology Market - Global Industry Analysis, Size, Share, Growth, Trends and Forecast 2015 - 2021 report at

Post has attachment
BMVA Workshop on Visual, tactile and force sensing for robot manipulation

One Day BMVA symposium in London, UK on Wednesday 9th December, 2015

Chairs: Lorenzo Jamone & Serena Ivaldi
Keynote speakers:  Kaspar Althoefer (King's College London, UK). Jeremy Wyatt (University of Birmingham, UK). Bruno Siciliano (University of Naples Federico II, Italy).
Book online at £10 for BMVA Members, £30 for Non Members  including lunch

The workshop will bring together researchers that are investigating data-driven strategies for learning and control of robotic manipulation, as well as specialists from academia and industry that are working on the development of novel sensing technologies that are relevant for grasping and manipulation, e.g. 2D and 3D vision, depth sensors, soft tactile sensors, compliant force sensing.
9.15: Welcome & intro
9.30: [Keynote 1] Bruno Siciliano (University of Naples) - Vision for robotic manipulation.
10.15: [Short talk 1] Jinguo Tian (University of Manchester) - Simulation of WireFrame Object Representations for a 3D Robotic Vision System.
10.30: Coffee break
11.00: [Short talk 2] Serena Ivaldi (INRIA Nancy) - Multimodal object learning with iCub.
11.15: [Keynote 2] Jeremy Wyatt (University of Birmingham) - Steps towards manipulation.
12.00: Lunch break
13.00: [Invited 1] Lorenzo Natale & Vadim Tikhanoff (IIT) - Seeing and touching objects: an overview of the iCub tactile and visual capabilities.
13.35: [Short talk 3] Zhe Su (University of Southern California) - Slip Classification Using Tangential and Torsional Skin Distortions on a Biomimetic Tactile Sensor.
13.50: [Invited 2] Antonio Morales (Universitat Jaume-I) - Sensor-based robust grasping and manipulation.
14.25: [Short talk 4] Lorenzo Jamone (University of Lisbon) - Learning sensorimotor model for robotic manipulation from human demonstration.
14.40: Coffee break
15.10: [Invited 3] Yasemin Bekiroglu (KTH) - Learning to assess grasp stability from vision, touch and proprioception.
15.45: [Keynote 3] Kaspar Althoefer (King's College London) - Combining vision and touch for object manipulation.
16.30: Conclusion & Discussion
17.00: Move to pub
Book online at £10 for BMVA Members, £30 for Non Members  including lunch

Post has attachment
CVMP 2015  - Registration Open

12th European Conference on Visual Media Production (CVMP)
British Film Institute (BFI), London, UK.  24-25 November 2015.

** Registration Now Open **

For over a decade CVMP has built a reputation as the prime venue for researchers to meet with practitioners in the Creative Industries: film, broadcast, games.  The Conference brings together expertise in video processing, computer vision, computer graphics, animation and physical simulation. It provides a forum for presentation of the latest research and application advances, combined with keynote and invited talks on state-of-the-art industry practice.

CVMP 2015, the 12th European Conference on Visual Media Production, is an ACM published conference.
:: KEYNOTES - Spanning Academic and Industry
* Lourdes Agapito (UCL) 
* Stuart Adcock (Industrial Light and Magic) 
* Oliver James (Double Negative Visual Effects) 
* Sameer Agarwal (Google / Jump Camera)
:: INVITED INDUSTRY TALKS  including Pixar and Autodesk
:: RESEARCH PAPERS: A vibrant technical papers and poster programme, available to view now at  
- 4 invited talks from VR industry specialists including The Third Floor, Happy Finish.
- 4 invited talks including the DREAMSPACE, REFRAME, GRIP and ASAP projects spanning academic and industry.
** Registration Now Open - Discount deals available for group registrations! **  
Come along to London, BFI 24-25 November to see the latest technology and dissemination of practice in visual media production / visual effects, for film, games, and broadcast.

Post has attachment

Post has attachment
BMVA Technical Meeting: The Computational Face – Automatic Face Analysis and Synthesis

Final Programme

One Day BMVA symposium in London, UK on Wednesday 14th October, 2015

Chairs: Brais Martinez, Yorgos Tzimiropoulos and Michel Valstar

Keynote speakers:  Tim Cootes (University of Manchester), Darren Cosker (University of Bath),  Maja Pantic (Imperial College London), Richard Bowden (University of Surrey),

In this meeting we hope to seek the opinions of speakers from the most interesting companies and academic labs with an interest in the computational face on the future of Face Analysis and Synthesis. What academic challenges remain? What elements are ready for adoption by industry? What are the key factors blocking progress? What new application opportunities can be envisaged?
09:30 Registration + Coffee
10:00 Keynote 1: Maja Pantic (Imperial College London)
10:40 Evangelos Sariyanidi (Queen Mary University London) - Probabilistic Subpixel Temporal Registration for Facial Expression Analysis
11:00 Anastasios Roussos (Imperial College London) - Dense variational 3D modelling and reconstruction of faces
11:20 Coffee Break
11:40 Keynote 2: Darren Cosker (University of Bath)  - Dynamic Facial Processing and Capture in Academia and Industry
12:20 Will Smith (University of York) - An example-based approach to face analysis and synthesis
12:40 Patrik Huber (Uni of Surrey) - Landmark Detection and 3D Face Reconstruction using Modern C++
13:00 Lunch Inc Posters:
Alessio Dore (Wirewax) - Face Detection, Tracking and Recognition for Interactive Videos
Ali Bukar (University of Bradford) - Nonlinear AAMs for age estimation and synthesis
Georgia Rajamanoharan (University of Nottingham) - Multi-View Constrained Local Models for Large Head Angle Facial Tracking
Joanna Olszewska (University of Gloucestershire) - Automated Face Detection: Challenges and Solutions

14:00 Keynote 3: Tim Cootes (University of Manchester) - Facial Feature Detection and Tracking using Statistical Shape Models
14:40 Yiannis Patras (Queen Mary University London) - Facial alignment in the wild - what does the mirror image tell us about the alignment error?
15:00 Thomas Heseltine (Aurora Computer Services) - Deep Learning for Biometrics in Real World Applications
15:20   Coffee Break 2
15:40   Keynote 4: Richard Bowden (University of Surrey)
16:20 Yi-Zhe Song (Queen Mary University London) - Cross-Modal Face Matching: Beyond Viewed Sketches
16:40 Charlie Frowd (University of Winchester) - Facial synthesis from human memory
17:00   End of meeting

Book online at £10 for BMVA Members
£30 for Non Members
Wait while more posts are being loaded