Iterative Text-Based Editing of Talking-Heads Using Neural Retargeting

We present a text-based tool for editing talking-head video that enables an iterative editing workflow. On each iteration users can edit the wording of the speech, further refine mouth motions if necessary to reduce artifacts, and manipulate non-verbal aspects of the performance by inserting mouth gestures (e.g., a smile) or changing the overall performance style (e.g., energetic, mumble). Our tool requires only 2 to 3 minutes of the target actor video and it synthesizes the video for each iteration in about 40 seconds, allowing users to quickly explore many editing possibilities as they iterate. Our approach is based on two key ideas. (1) We develop a fast phoneme search algorithm that can quickly identify phoneme-level subsequences of the source repository video that best match a desired edit. This enables our fast iteration loop. (2) We leverage a large repository of video of a source actor and develop a new self-supervised neural retargeting technique for transferring the mouth motions of the source actor to the target actor. This allows us to work with relatively short target actor videos, making our approach applicable in many real-world editing scenarios. Finally, our, refinement and performance controls give users the ability to further fine-tune the synthesized results.

Who Is Okay? The Harm of One-Dimensional Appraisals of Women Scholars During COVID-19 & Beyond

In this essay, I detail how homogenizing appraisals of diverse faculty women during COVID-19 are harmful to all, including myself. I highlight how academic demands to be “talking heads” and not full human beings, though not new, is especially harmful in the current era. As a Black woman faculty dealing with the double pandemic of COVID-19 and anti-Black racism, the one-dimensional appraisals of women faculty exclude me: I am not a mother dealing with sexist overburden in household responsibilities that interfere with my work. Instead, I am dealing with isolation and loneliness, which I sublimate through work productivity. Resulting in shame, I also realize that universities could operate differently, recognizing women scholars for their diversity in identities, backgrounds, responsibilities, work styles, and personalities during the pandemic and beyond. Given that work productivity is not synonymous with well-being, I hope my colleagues know that, in this moment, I am not okay.

Talking heads and shitting in the street: Stuttering Parrhesia in three modes

This paper seeks both to expand the range of what counts as political action for dysfluent voices and to find resources that can generate critical breaks within neo-liberal modes of power. With the Cynics, I suggest that some truths – like dysfluent lives are worth living – cannot be told by a talking head. I accordingly map three possible modes of truth-telling within the lexicon of parrhesia: therapeutic, Platonic and Cynic. Therapeutic truth-telling is an apolitical enunciation that indexes a model of authenticity and is limited to speaking truth about oneself and the world in a normalizing register. Platonic parrhesia is a form of equality-based political discourse that aims at inclusion. In this mode, the parrhesiastes, like the talking head, must fashion their body as a pure vessel of truth to be recognized as such. Cynic truth-telling, finally, is a radical embodiment of critique that seeks rupture rather than understanding. Taking up the motto of the Cynics – ‘deface the currency’ – perhaps dysfluent voices can find resources to ‘de-face’ speech and its mythic power that has become entwined with capital.