Valerio Giuffrida

valerio giuffrida

Dr Valerio Giuffrida

  

Biography

Valerio obtained his PhD in Image Analysis at IMT School for Advanced Studies Lucca (Italy) in the 2018, with a thesis entitled “Learning to Count Leaves of Plants”. He has been engaged in several projects related to non-destructive image-based plant phenotyping. He proposed the first award-winning machine learning approach to predict the number of leaves of rosette plants, presented at Computer Vision Problems in Plant Phenotyping Workshop held in conjunction with the BMVC 2015. This algorithm was also included in the Phenotiki Analysis Software (http://phenotiki.com), software also developed in collaboration with him and has been downloaded more than 100 times. His expertise includes deep learning, publishing several papers related to deep leaf counting [27], [30], plant image synthesis [31], and domain adaptation applied plant phenotyping [4]. Several of these works were the first real applications in plant phenotyping. He also collaborated for the BBSRC project BB/P023487/1 as PDRA at the University of Edinburgh, for which he developed the Phenotiki Sensors and the analysis algorithms to extract root system architecture (RSA) traits.

His research interests are computer vision, machine learning, and deep learning applied to plant phenotyping and medical imaging applications. He also focuses in explainability and interpretability of deep networks, and transfer learning (domain adaptation).

Research Areas

Esteem

Invited Speaker

  • Computer Vision for Agriculture
  • Camera Trap Symposium
  • Big Data in Society Meeting
  • Presentation of the paper 'Explicit Factorization of Rotations in Restricted Boltzmann Machines'
  • Generating Images using Neural Networks

 

Date


16 results

Towards Continuous User Authentication Using Personalised Touch-Based Behaviour

Conference Proceeding
Aaby, P., Giuffrida, M. V., Buchanan, W. J., & Tan, Z. (in press)
Towards Continuous User Authentication Using Personalised Touch-Based Behaviour
In this paper, we present an empirical evaluation of 30 features used in touch-based continuous authentication. It is essential to identify the most significant features for e...

Affordable and robust phenotyping framework to analyse root system architecture of soil-grown plants

Journal Article
Bontpart, T., Concha, C., Giuffrida, V., Robertson, I., Admkie, K., Degefu, T., …Doerner, P. (in press)
Affordable and robust phenotyping framework to analyse root system architecture of soil-grown plants. Plant Journal, https://doi.org/10.1111/tpj.14877
The phenotypic analysis of root system growth is important to inform efforts to enhance plant resource acquisition from soils. However, root phenotyping still remains challeng...

Doing More With Less: A Multitask Deep Learning Approach in Plant Phenotyping

Journal Article
Dobrescu, A., Giuffrida, M. V., & Tsaftaris, S. A. (2020)
Doing More With Less: A Multitask Deep Learning Approach in Plant Phenotyping. Frontiers in Plant Science, 11, https://doi.org/10.3389/fpls.2020.00141
Image-based plant phenotyping has been steadily growing and this has steeply increased the need for more efficient image analysis techniques capable of evaluating multiple pla...

Unsupervised Rotation Factorization in Restricted Boltzmann Machines

Journal Article
Giuffrida, M. V., & Tsaftaris, S. A. (2020)
Unsupervised Rotation Factorization in Restricted Boltzmann Machines. IEEE Transactions on Image Processing, 29(1), 2166-2175. https://doi.org/10.1109/TIP.2019.2946455
Finding suitable image representations for the task at hand is critical in computer vision. Different approaches extending the original Restricted Boltzmann Machine (RBM) mode...

Leaf Counting Without Annotations Using Adversarial Unsupervised Domain Adaptation

Conference Proceeding
Giuffrida, M. V., Dobrescu, A., Doerner, P., & Tsaftaris, S. A. (2019)
Leaf Counting Without Annotations Using Adversarial Unsupervised Domain Adaptation
Deep learning is making strides in plant phenotyping and agriculture. But pretrained models require significant adaptation to work on new target datasets originating from a di...

Adversarial Large-scale Root Gap Inpainting

Conference Proceeding
Chen, H., Giuffrida, M. V., Doerner, P., & Tsaftaris, S. A. (2019)
Adversarial Large-scale Root Gap Inpainting
Root imaging of a growing plant in a non-invasive, affordable , and effective way remains challenging. One approach is to image roots by growing them in a rhizobox, a soil-fil...

Root Gap Correction with a Deep Inpainting Model

Conference Proceeding
Chen, H., Giuffrida, M. V., Doerner, P., & Tsaftaris, S. A. (2018)
Root Gap Correction with a Deep Inpainting Model
Imaging roots of growing plants in a non-invasive and affordable fashion has been a long-standing problem in image-assisted plant breeding and phenotyping. One of the most aff...

Pheno-Deep Counter: a unified and versatile deep learning architecture for leaf counting

Journal Article
Giuffrida, M. V., Doerner, P., & Tsaftaris, S. A. (2018)
Pheno-Deep Counter: a unified and versatile deep learning architecture for leaf counting. Plant Journal, 96(4), 880-890. https://doi.org/10.1111/tpj.14064
Direct observation of morphological plant traits is tedious and a bottleneck for high‐throughput phenotyping. Hence, interest in image‐based analysis is increasing, with the r...

Citizen crowds and experts: observer variability in image-based plant phenotyping

Journal Article
Giuffrida, M. V., Chen, F., Scharr, H., & Tsaftaris, S. A. (2018)
Citizen crowds and experts: observer variability in image-based plant phenotyping. Plant Methods, 14(1), https://doi.org/10.1186/s13007-018-0278-7
Background: Image-based plant phenotyping has become a powerful tool in unravelling genotype–environment interactions. The utilization of image analysis and machine learning h...

Leveraging multiple datasets for deep leaf counting

Conference Proceeding
Dobrescu, A., Giuffrida, M. V., & Tsaftaris, S. A. (2018)
Leveraging multiple datasets for deep leaf counting. https://doi.org/10.1109/ICCVW.2017.243
The number of leaves a plant has is one of the key traits (phenotypes) describing its development and growth. Here, we propose an automated, deep learning based approach for c...

Current Post Grad projects