Multivariate time series (MTS) play essential roles in daily life because most real-world time series datasets are multivariate and rich in time-dependent information. Traditional forecasting methods for MTS are time-consuming and filled with complicated limitations. One efficient method being explored within the dynamical systems is the extended short-term memory networks (LSTMs). However, existing MTS models only partially use the hidden spatial relationship as effectively as LSTMs. Shallow LSTMs are inadequate in extracting features from high-dimensional MTS; however, the multilayer bidirectional LSTM (BiLSTM) can learn more MTS features in both directions. This study tries to generate a novel and improved BiLSTM network (DBI-BiLSTM) based on a deep belief network (DBN), bidirectional propagation technique, and a chained structure. The deep structures are constructed by a DBN layer and multiple stacked BiLSTM layers, which increase the feature representation of DBI-BiLSTM and allow for the model to further learn the extended features in two directions. First, the input is processed by DBN to obtain comprehensive features. Then, the known features, divided into clusters based on a global sensitivity analysis method, are used as the inputs of every BiLSTM layer. Meanwhile, the previous outputs of the shallow layer are combined with the clustered features to reconstitute new input signals for the next deep layer. Four experimental real-world time series datasets illustrate our one-step-ahead prediction performance. The simulating results confirm that the DBI-BiLSTM not only outperforms the traditional shallow artificial neural networks (ANNs), deep LSTMs, and some recently improved LSTMs, but also learns more features of the MTS data. As compared with conventional LSTM, the percentage improvement of DBI-BiLSTM on the four MTS datasets is 85.41, 75.47, 61.66 and 30.72%, respectively.

Surround modulation has been abundantly studied in several mammalian brain areas, including the primary visual cortex, lateral geniculate nucleus, and superior colliculus (SC), but systematic analysis is lacking in the avian optic tectum (OT, homologous to mammal SC). Here, multi-units were recorded from pigeon (Columba livia) OT, and responses to different sizes of moving, flashed squares, and bars were compared. The statistical results showed that most tectal neurons presented suppressed responses to larger stimuli in both moving and flashed paradigms, and suppression induced by flashed squares was comparable with moving ones when the stimuli center crossed the near classical receptive field (CRF) center, which corresponded to the full surrounding condition. Correspondingly, the suppression grew weaker when the stimuli center moved across the CRF border, equivalent to partially surrounding conditions. Similarly, suppression induced by full surrounding flashed squares was more intense than by partially surrounding flashed bars. These results suggest that inhibitions performed on tectal neurons appear to be full surrounding rather than locally lateral. This study enriches the understanding of surround modulation properties of avian tectum neurons and provides possible hypotheses about the arrangement of inhibitions from other nuclei, both of which are important for clarifying the mechanism of target detection against clutter background performed by avians.

Birds possess remarkable visual object classification abilities. Objects in nature often contain multiple redundant cues (e.g., color and shape). Previous work has shown that pigeons prefer color over shape for target classification, but it remains unknown whether they rely primary on color or whether shape also provides discriminative information when both cues are available. To address this, we trained pigeons on single-cue (color-only or shape-only) discrimination tasks and then transferred them to multi-cue (color + shape) tasks, while simultaneously recording neural responses from the entopallium. Behaviorally, shifting from single-color to multi-cue produced no change in recognition accuracy, whereas shifting from single-shape to multi-cue significantly improved accuracy. This pattern suggests that color provided the primary behavioral driver, with shape adding limited detectable benefit under these specific conditions. Neurophysiologically, most recording sites in the entopallial showed stronger local field potential responses to color than to shape cues, and local functional connectivity was less dense under multi-cue than single-cue images. Together, these findings suggest that, under the present redundant-cue paradigm, pigeons exhibited a preferential reliance on color over shape for object recognition a processing bias that may reflect an efficient strategy when one cue is sufficient for accurate classification. These findings contribute to our understanding of how birds process redundant visual information under conditions of cue redundancy.

Understanding how birds perceive and recognize objects remains a fundamental question in neuroscience. The entopallium, a key node in the avian tectofugal pathway, is involved in complex visual processing, yet its functional organization remains incompletely understood. Here, we systematically mapped entopallium neuronal responses in pigeons using controlled visual stimuli varying in color, shape, and motion. Entopallium neurons exhibited strong hue selectivity invariant to luminance, pronounced orientation tuning to shape stimuli, and robust direction selectivity to motion. Spatial analyses revealed distinct functional segregation, with color-selective neurons localized anteroventrally, shape-selective neurons dorsally, and motionselective neurons posteriorly. Despite this segregation, a subset of neurons exhibited joint tuning across stimulus dimensions, suggesting an integrated yet modular organizational scheme. Notably, entopallium neurons exhibited a moderate level of feature integration, sharing key characteristics with early to intermediate stages of mammalian visual processing. Together, these findings establish the entopallium as a central hub for multidimensional visual processing in birds and provide new insights into the convergent evolution of sophisticated visual systems across vertebrates.

[This corrects the article DOI: 10.1007/s11571-022-09916-w.].