Automatic metrics are widely used in machine translation as a substitute for human assessment. With the introduction of any new metric comes the question of just how well that metric mimics human assessment of translation quality. This is often measured by correlation with human judgment. Significance tests are generally not used to establish whether improvements over existing methods such as B...

The tracking of humans in a video stream has become one of the most desirable computer vision tasks over the past few years. It remains however a difficult problem and the reliability of systems is often dependent on getting good lighting and clear video images. This paper reports on the development of our PHACT tracker: parallel HOG and correlation tracking. This system uses a cascade of track...

The task of automatically generating sentential descriptions of image content has become increasingly popular in recent years, resulting in the development of large-scale image description datasets and the proposal of various metrics for evaluating image description generation systems. However, not much work has been done to analyse and understand both datasets and the metrics. In this paper, w...

Automatic evaluation metrics are fundamentally important for Machine Translation, allowing comparison of systems performance and efficient training. Current evaluation metrics fall into two classes: heuristic approaches, like BLEU, and those using supervised learning trained on human judgement data. While many trained metrics provide a better match against human judgements, this comes at the co...

Automatic evaluation metrics are fundamentally important for Machine Translation, allowing comparison of systems performance and efficient training. Current evaluation metrics fall into two classes: heuristic approaches, like BLEU, and those using supervised learning trained on human judgement data. While many trained metrics provide a better match against human judgements, this comes at the co...

Metrics for human-robot teaming should extend to teams consisting of multiple human and robotic agents, working in complex, dynamic work domains. This proposes that comprehensively analyze evaluate multi-human, multi-robot teams, traditional HRI metrics performance, efficiency must be expanded upon incorporate teamwork. We develop five distinct capture both ecological cognitive aspects teamwork...

This study presents a new model of the human image quality assessment process: the aim is to highlight the foundations of the image quality metrics proposed in literature, by identifying the cognitive/physiological or mathematical principles of their development and the relation with the actual human quality assessment process. The model allows to create a novel categorization of objective and ...

A procedure to select the most relevant metrics for assessing the ecological condition of the Douro basin (north Portugal) was developed based upon a set of 184 benthic community metrics. They were grouped into 16 biological categories selected from literature using data collected over 2 years from 54 sites along 31 rivers covering the whole perceived range of human disturbance. Multivariate an...