scholarly journals Bimodal latency distribution and distractor effects in Express Saccades in humans.

2016 ◽  
Vol 16 (12) ◽  
pp. 854
Author(s):  
Jessica Heeman ◽  
Stefan Van der Stigchel ◽  
Jan Theeuwes
Keyword(s):  
Latency Distribution ◽  
Express Saccades

scholarly journals Air Learning: a deep reinforcement learning gym for autonomous aerial robot visual navigation

2021 ◽  
Author(s):  
Srivatsan Krishnan ◽  
Behzad Boroujerdian ◽  
William Fu ◽  
Aleksandra Faust ◽  
Vijay Janapa Reddi
Keyword(s):  
Reinforcement Learning ◽  
Embedded System ◽  
Broad Class ◽  
Visual Navigation ◽  
Raspberry Pi ◽  
Latency Distribution ◽  
Hardware In The Loop ◽  
Resource Constrained ◽  
Aerial Robot ◽  
Policy Optimization

AbstractWe introduce Air Learning, an open-source simulator, and a gym environment for deep reinforcement learning research on resource-constrained aerial robots. Equipped with domain randomization, Air Learning exposes a UAV agent to a diverse set of challenging scenarios. We seed the toolset with point-to-point obstacle avoidance tasks in three different environments and Deep Q Networks (DQN) and Proximal Policy Optimization (PPO) trainers. Air Learning assesses the policies’ performance under various quality-of-flight (QoF) metrics, such as the energy consumed, endurance, and the average trajectory length, on resource-constrained embedded platforms like a Raspberry Pi. We find that the trajectories on an embedded Ras-Pi are vastly different from those predicted on a high-end desktop system, resulting in up to $$40\%$$ 40 % longer trajectories in one of the environments. To understand the source of such discrepancies, we use Air Learning to artificially degrade high-end desktop performance to mimic what happens on a low-end embedded system. We then propose a mitigation technique that uses the hardware-in-the-loop to determine the latency distribution of running the policy on the target platform (onboard compute on aerial robot). A randomly sampled latency from the latency distribution is then added as an artificial delay within the training loop. Training the policy with artificial delays allows us to minimize the hardware gap (discrepancy in the flight time metric reduced from 37.73% to 0.5%). Thus, Air Learning with hardware-in-the-loop characterizes those differences and exposes how the onboard compute’s choice affects the aerial robot’s performance. We also conduct reliability studies to assess the effect of sensor failures on the learned policies. All put together, Air Learning enables a broad class of deep RL research on UAVs. The source code is available at: https://github.com/harvard-edge/AirLearning.

An examination of the variables that affect express saccade generation

2004 ◽  
Vol 21 (2) ◽  
pp. 119-127 ◽  
Author(s):  
PETER H. SCHILLER ◽  
JOHANNES HAUSHOFER ◽  
GEOFFERY KENDALL
Keyword(s):  
Rhesus Monkeys ◽  
Target Onset ◽  
Express Saccades ◽  
Express Saccade ◽  
Saccadic System ◽  
Fixation Spot ◽  
Temporal Asynchrony ◽  
Saccade Generation ◽  
Gap Time

The frequency with which express saccades are generated under a variety of conditions in rhesus monkeys was examined. Increasing the gap time between fixation spot termination and target onset increased express saccade frequency but was progressively less effective in doing so as the number of target positions in the sample was increased. Express saccades were rarely produced when two targets were presented simultaneously and the choice of either of which was rewarded; a temporal asynchrony of only 17 ms between the targets reinstated express saccade generation. Express saccades continued to be generated when the vergence or pursuit systems was coactivated with the saccadic system.

scholarly journals Parietal Cortical Neuronal Activity Is Selective for Express Saccades

2013 ◽  
Vol 33 (2) ◽  
pp. 814-823 ◽  
Author(s):  
M. Chen ◽  
Y. Liu ◽  
L. Wei ◽  
M. Zhang
Keyword(s):  
Neuronal Activity ◽  
Express Saccades

Saccadic Eye Movements of Dyslexic and Normal Reading Children

Perception ◽  
1994 ◽  
Vol 23 (1) ◽  
pp. 45-64 ◽  
Author(s):  
Monica Biscaldi ◽  
Burkhart Fischer ◽  
Franz Aiple
Keyword(s):  
Eye Movements ◽  
Reaction Times ◽  
Saccadic Eye Movements ◽  
Control Group ◽  
Express Saccades ◽  
Movement Data ◽  
Fixation Durations ◽  
Normal Reading ◽  
Single Target ◽  
The Right

Twenty-four children made saccades in five noncognitive tasks. Two standard tasks required saccades to a single target presented randomly 4 deg to the right or left of a fixation point. Three other tasks required sequential saccades from the left to the right. 75 parameters of the eye-movement data were collected for each child. On the basis of their reading, writing, and other cognitive performances, twelve children were considered dyslexic and were divided into two groups (D1 and D2). Group statistical comparisons revealed significant differences between control and dyslexic subjects. In general, in the standard tasks the dyslexic subjects had poorer fixation quality, failed more often to hit the target at once, had smaller primary saccades, and had shorter reaction times to the left as compared with the control group. The control group and group D1 dyslexics showed an asymmetrical distribution of reaction times, but in opposite directions. Group D2 dyslexics made more anticipatory and express saccades, they undershot the target more often in comparison with the control group, and almost never overshot it. In the sequential tasks group D1 subjects made fewer and larger saccades in a shorter time and group D2 subjects had shorter fixation durations than the subjects of the control group.

The effect of frontal eye field and superior colliculus lesions on saccadic latencies in the rhesus monkey

1987 ◽  
Vol 57 (4) ◽  
pp. 1033-1049 ◽  
Author(s):  
P. H. Schiller ◽  
J. H. Sandell ◽  
J. H. Maunsell
Keyword(s):  
Eye Movements ◽  
Superior Colliculus ◽  
Short Latency ◽  
The Other ◽  
Frontal Eye Field ◽  
Express Saccades ◽  
Unimodal Distribution ◽  
Other Hand ◽  
Eye Field ◽  
The Mean

Rhesus monkeys were trained to make saccadic eye movements to visual targets using detection and discrimination paradigms in which they were required to make a saccade either to a solitary stimulus (detection) or to that same stimulus when it appeared simultaneously with several other stimuli (discrimination). The detection paradigm yielded a bimodal distribution of saccadic latencies with the faster mode peaking around 100 ms (express saccades); the introduction of a pause between the termination of the fixation spot and the onset of the target (gap) increased the frequency of express saccades. The discrimination paradigm, on the other hand, yielded only a unimodal distribution of latencies even when a gap was introduced, and there was no evidence for short-latency "express" saccades. In three monkeys either the frontal eye field or the superior colliculus was ablated unilaterally. Frontal eye field ablation had no discernible long-term effects on the distribution of saccadic latencies in either the detection or discrimination tasks. After unilateral collicular ablation, on the other hand, express saccades obtained in the detection paradigm were eliminated for eye movements contralateral to the lesion, leaving only a unimodal distribution of latencies. This deficit persisted throughout testing, which in one monkey continued for 9 mo. Express saccades were not observed again for saccades contralateral to the lesion, and the mean latency of the contralateral saccades was longer than the mean latency of the second peak for the ipsiversive saccades. The latency distribution of saccades ipsiversive to the collicular lesion was unaffected except for a few days after surgery, during which time an increase in the proportion of express saccades was evident. Saccades obtained with the discrimination paradigm yielded a small but reliable increase in saccadic latencies following collicular lesions, without altering the shape of the distribution. Unilateral muscimol injections into the superior colliculus produced results similar to those obtained immediately after collicular lesions: saccades contralateral to the injection site were strongly inhibited and showed increased saccadic latencies. This was accompanied by a decrease of ipsilateral saccadic latencies and an increase in the number of saccades falling into the express range. The results suggest that the superior colliculus is essential for the generation of short-latency (express) saccades and that the frontal eye fields do not play a significant role in shaping the distribution of saccadic latencies in the paradigms used in this study.(ABSTRACT TRUNCATED AT 400 WORDS)

scholarly journals Express saccades elicited during visual scan in the monkey

1994 ◽  
Vol 34 (15) ◽  
pp. 2023-2038 ◽  
Author(s):  
M.A. Sommer
Keyword(s):  
Express Saccades
Sign in / Sign up

Export Citation Format

Share Document