Why does co-speech gesture recognition using convolutional neural networks often fail to generalize across diverse communication styles in a multimodal interface?

💡 Explanation

Co-speech gesture recognition accuracy declines because variability in gesture iconicity (the degree to which gestures resemble their referents) affects feature extraction; therefore, the model struggles to learn consistent mappings, rather than purely spatial or temporal feature limitations.

• Why does text using a condensed font with tight inter-character spacing significantly reduce readability on e-ink displays under low lighting conditions?
• Why does a researcher publishing outside their expertise risk decreased credibility?
• Why does the presence of a strong accent typically lead to a broadened acoustic spectrum during speech processing, affecting automatic speech recognition (ASR) accuracy?
• Why does color perception differ across languages, even given identical stimuli?
• Why does speech production under cognitive load rely increasingly on formulaic language?
• A cognitive scientist analyzes eye-tracking data from participants reading a novel; if participants exhibit longer fixation durations on specific words, what explains this phenomenon in eye movements?