Conference paper Open Access

Comparing CNNs and JPEG for Real-Time Multi-view Streaming in Tele-Immersive Scenarios

Konstantoudakis, Konstantinos; Christakis, Emmanouil; Drakoulis, Petros; Doumanoglou, Alexandros; Zioulis, Nikolaos; Zarpalas, Dimitrios; Daras, Petros


Citation Style Language JSON Export

{
  "DOI": "10.1109/SITIS.2018.00022", 
  "language": "eng", 
  "title": "Comparing CNNs and JPEG for Real-Time Multi-view Streaming in Tele-Immersive Scenarios", 
  "issued": {
    "date-parts": [
      [
        2018, 
        11, 
        26
      ]
    ]
  }, 
  "abstract": "<p>Deep learning-based codecs for lossy image compression have recently managed to surpass traditional codecs like JPEG and JPEG 2000 in terms of rate-distortion tradeoff. However, they generally utilize architectures with large numbers of stacked layers, often making their inference execution prohibitively slow for time-sensitive applications. In this work, we assess the suitability of such compression techniques in real-time video streaming, and, more specifically, next-generation interactive tele-presence applications, which impose stringent latency requirements. To that end, we compare a recently published work on image compression based on convolutional neural networks which achieves state-of-the-art compression ratio using a relatively lightweight architecture, against a CPU and a GPU implementation of JPEG, measuring compression ratios and timings. With these results, we run a simulation of a tele-immersion pipeline for various networking conditions and examine the performance of the compared codecs, calculating framerates and latencies for different codec/network combinations.</p>", 
  "author": [
    {
      "family": "Konstantoudakis, Konstantinos"
    }, 
    {
      "family": "Christakis, Emmanouil"
    }, 
    {
      "family": "Drakoulis, Petros"
    }, 
    {
      "family": "Doumanoglou, Alexandros"
    }, 
    {
      "family": "Zioulis, Nikolaos"
    }, 
    {
      "family": "Zarpalas, Dimitrios"
    }, 
    {
      "family": "Daras, Petros"
    }
  ], 
  "type": "paper-conference", 
  "id": "3137854"
}
26
25
views
downloads
Views 26
Downloads 25
Data volume 31.8 MB
Unique views 25
Unique downloads 22

Share

Cite as