{
  "absUrl": "https://arxiv.org/abs/2610.36834v1",
  "arxivId": "2610.36834",
  "arxivVersion": 1,
  "arxivVersionedId": "2610.36834v1",
  "authors": [
    {
      "affiliations": [
        "School of Computer Science, Wuhan University, Wuhan, China",
        "School of Artificial Intelligence, Chinese University of Hong Kong, Shenzhen, China"
      ],
      "name": "Zhan Jin"
    },
    {
      "affiliations": [
        "School of Artificial Intelligence, Wuhan University, Wuhan, China",
        "School of Cyber Science and Engineering, Wuhan University, Wuhan, China"
      ],
      "name": "Peijun Yang"
    },
    {
      "affiliations": [
        "School of Artificial Intelligence, Wuhan University, Wuhan, China",
        "School of Cyber Science and Engineering, Wuhan University, Wuhan, China"
      ],
      "name": "Juan Liu"
    },
    {
      "affiliations": [
        "School of Artificial Intelligence, Chinese University of Hong Kong, Shenzhen, China",
        "School of Artificial Intelligence, Wuhan University, Wuhan, China"
      ],
      "name": "Ming Li"
    }
  ],
  "contract": "researcher-sidecars-v1",
  "id": "2610.36834v1",
  "originalTitle": "A Lightweight Dual-Branch Visual Front-End for Efficient Audio-Visual Target Speaker Extraction",
  "pdfUrl": "https://arxiv.org/pdf/2610.36834v1.pdf",
  "schemaVersion": 1,
  "type": "preprint"
}