Memories are One-to-Many Mapping Alleviators in Talking Face Generation.

Talking face generation aims at generating photorealistic video portraits of a target person driven by input audio. According to the nature of audio to lip motions mapping, the same speech content may have different appearances even for the same person at different occasions. Such one-to-many mapping problem brings ambiguity during training and thus causes inferior visual results. Although this one-to-many mapping could be alleviated in part by a two-stage framework (i.e., an audioto- expression model followed by a neural-rendering model), it is still insufficient since the prediction is produced without enough information (e.g., emotions, wrinkles, etc.). In this paper, we propose MemFace to complement the missing information with an implicit memory and an explicit memory that follow the sense of the two stages respectively. More specifically, the implicit memory is employed in the audio-to-expression model to capture high-level semantics in the audio-expression shared space, while the explicit memory is employed in the neural-rendering model to help synthesize pixel-level details. Our experimental results show that our proposed MemFace surpasses all the state-of-theart results across multiple scenarios consistently and significantly.

(Bio)dissolution of arsenopyrite coupled with multiple proportions of pyrite: Emphasis on the mobilization and existential state of arsenic.

The formation of arsenic-bearing acid mine drainage (AMD) via the oxidation of arsenopyrite refuse ore has attracted significant attention. Pyrite, as main a concomitant mineral, is a crucial factor that affects the (bio)dissolution of arsenopyrite, but there are still some points on the detailed action mechanism under normal environmental conditions that need further study. In this study, the effect mechanism of pyrite with a systematic pyrite content (0, 10, 25, 50, 75, 90, and 100 wt %) on arsenopyrite oxidation and arsenic release in the presence of Acidithiobacillus ferrooxidans was investigated. The X-ray diffraction (XRD), scanning election microscopy (SEM), X-ray photoelectron spectroscopy (XPS) and electrochemical analyses were also carried out. Results showed that the existence of pyrite and Acidithiobacillus ferrooxidans significantly accelerated the dissolution of arsenopyrite and the oxidation of As (Ⅲ) to As (Ⅴ), resulting from the galvanic effect, an increase in the Fe3+/Fe2+ ratio and the oxidation-reduction potential (Eh) value, and a decrease in pH level. As the detected main intermediate products, element sulphur was considered as the dominating obstructive factor during arsenopyrite oxidation, while the added pyrite could accelerate its oxidation. Moreover, a close relationship between different mineral proportions and the galvanic effect was also observed and discussed. Finally, suggestions on AMD governance and source control are proposed.