Key points Vision plays a crucial role in guiding locomotion in complex environments, but the coordination between gaze and stride is not well understood. The coordination of gaze shifts, fixations, constant gaze and slow gaze with strides in cats walking on different surfaces were examined. It was found that gaze behaviours are coordinated with strides even when walking on a flat surface in the complete darkness, occurring in a sequential order during different phases of the stride. During walking on complex surfaces, gaze behaviours are typically more tightly coordinated with strides, particularly at faster speeds, only slightly shifting in phase. These findings indicate that the coordination of gaze behaviours with strides is not vision‐driven, but is a part of the whole body locomotion synergy; the visual environment and locomotor task modulate it. The results may be relevant to developing diagnostic tools and rehabilitation approaches for patients with locomotor deficits. Abstract Vision plays a crucial role in guiding locomotion in complex environments. However, the coordination between the gaze and stride is not well understood. We investigated this coordination in cats walking on a flat surface in darkness or light, along a horizontal ladder and on a pathway with small stones. We recorded vertical and horizontal eye movements and 3‐D head movement, and calculated where gaze intersected the walkway. The coordination of gaze shifts away from the animal, gaze shifts toward, fixations, constant gaze, and slow gaze with strides was investigated. We found that even during walking on the flat surface in the darkness, all gaze behaviours were coordinated with strides. Gaze shifts and slow gaze toward started in the beginning of each forelimb's swing and ended in its second half. Fixations peaked throughout the beginning and middle of swing. Gaze shifts away began throughout the second half of swing of each forelimb and ended when both forelimbs were in stance. Constant gaze and slow gaze away occurred in the beginning of stance. However, not every behaviour occurred during every stride. Light had a small effect. The ladder and stones typically increased the coordination and caused gaze behaviours to occur 3% earlier in the cycle. At faster speeds, the coordination was often tighter and some gaze behaviours occurred 2–16% later in the cycle. The findings indicate that the coordination of gaze with strides is not vision‐driven, but is a part of the whole body locomotion synergy; the visual environment and locomotor task modulate it.

Avoiding obstacles is essential for successful navigation through complex environments. This study aimed to clarify what strategies are used by a typical quadruped, the cat, to avoid obstacles during walking. Four cats walked along a corridor 2.5 m long and 25 or 15 cm wide. Obstacles, small round objects 2.5 cm in diameter and 1 cm in height, were placed on the floor in various locations. Movements of the paw were recorded with a motion capture and analysis system (Visualeyez, PTI). During walking in the wide corridor, cats' preferred strategy for avoiding a single obstacle was circumvention, during which the stride direction changed while stride duration and swing-to-stride duration ratio were preserved. Another strategy, stepping over the obstacle, was used during walking in the narrow corridor, when lateral deviations of walking trajectory were restricted. Stepping over the obstacle involved changes in two consecutive strides. The stride preceding the obstacle was shortened, and swing-to-stride ratio was reduced. The obstacle was negotiated in the next stride of increased height and normal duration and swing-to-stride ratio. During walking on a surface with multiple obstacles, both strategies were used. To avoid contact with the obstacle, cats placed the paw away from the object at a distance roughly equal to the diameter of the paw. During obstacle avoidance cats prefer to alter muscle activities without altering the locomotor rhythm. We hypothesize that a choice of the strategy for obstacle avoidance is determined by minimizing the complexity of neuro-motor processes required to achieve the behavioral goal. In a study of feline locomotor behavior we found that the preferred strategy to avoid a small obstacle is circumvention. During circumvention, stride direction changes but length and temporal structure are preserved. Another strategy, stepping over the obstacle, is used in narrow walkways. During overstepping, two strides adjust. A stride preceding the obstacle decreases in length and duration. The following stride negotiating the obstacle increases in height while retaining normal temporal structure and nearly normal length.

No abstract

No abstract

Speech intelligibility in cochlear implant (CI) users degrades considerably in listening environments with reverberation and noise. Previous research in automatic speech recognition (ASR) has shown that phoneme-based speech enhancement algorithms improve ASR system performance in reverberant environments as compared to a global model. However, phoneme-specific speech processing has not yet been implemented in CIs. In this paper, we propose a causal deep learning framework for classifying phonemes using features extracted at the time-frequency resolution of a CI processor. We trained and tested long short-term memory networks to classify phonemes and manner of articulation in anechoic and reverberant conditions. The results showed that CI-inspired features provide slightly higher levels of performance than traditional ASR features. To the best of our knowledge, this study is the first to provide a classification framework with the potential to categorize phonetic units in real-time in a CI.

In listening environments with room reverberation and background noise, cochlear implant (CI) users experience substantial difficulties in understanding speech. Because everyday environments have different combinations of reverberation and noise, there is a need to develop algorithms that can mitigate both effects to improve speech intelligibility. Desmond et al. (2014) developed a machine learning approach to mitigate the adverse effects of late reverberant reflections of speech signals by using a classifier to detect and remove affected segments in CI pulse trains. In this study, we investigate the robustness of the reverberation mitigation algorithm in environments with both reverberation and noise. We conducted sentence recognition tests in normal hearing listeners using vocoded speech with unmitigated and mitigated reverberant-only or noisy reverberant speech signals, across different reverberation times and noise types. Improvements in speech intelligibility were observed in mitigated reverberant-only conditions. However, mixed results were obtained in the mitigated noisy reverberant conditions as a reduction in speech intelligibility was observed for noise types whose spectra were similar to that of anechoic speech. Based on these results, the focus of future work will be to develop a context-dependent approach that activates different mitigation strategies for different acoustic environments. [Research supported by NIH grant R01DC014290-03.]

No abstract