Olona-Nkan Titele adanwo

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

1. Kini ibi-afẹde pataki ti ipasẹ ohun-ọpọlọpọ?
   • Lati tẹle awọn nkan pupọ kọja awọn fireemu lakoko titọju idanimọ deede fun ọkọọkan
   • Lati ṣe lẹtọ aworan kan
   • Lati compress fidio kan
   • Lati se ina titun awọn fireemu
2. Kini ilana 'titele-nipasẹ-iwari' pẹlu?
   • Ipasẹ laisi aṣawari eyikeyi
   • Ṣiṣawari awọn nkan ni fireemu kọọkan, lẹhinna sisopo awọn awari wọnyẹn kọja awọn fireemu
   • Lilo awọn ifẹnukonu ohun nikan
   • Ipasẹ nipasẹ awọn apoti iyaworan pẹlu ọwọ
3. Ipa wo ni àlẹmọ Kalman ṣe ninu awọn olutọpa bii SORT?
   • O compresses awọn aworan
   • O ṣe awari awọn nkan titun lati ibere
   • O ṣe asọtẹlẹ ibi ti ohun kọọkan tọpa yoo gbe ni atẹle
   • O ṣe afikun awọ si awọn fireemu
4. Kini DeepSORT ṣafikun lori oke SORT?
   • Iwọn aworan ti o tobi julọ
   • Titele ohun
   • Awọn atunkọ aifọwọyi
   • Ifisinu ifarahan ti o kọ ẹkọ fun tun-idamo awọn nkan lẹhin occlusion
5. Kini 'iyipada idanimọ' ni ipasẹ ohun-ọpọlọpọ?
   • Nigbati olutọpa ba paarọ awọn ID ti awọn nkan meji, nigbagbogbo bi wọn ṣe kọja awọn ọna
   • Yiyipada ipinnu fidio
   • Yipada kamẹra si pipa
   • Lorukọmii faili fidio naa
6. Bawo ni ByteTrack ṣe ilọsiwaju deede ipasẹ?
   • Nipa aibikita gbogbo awọn awari
   • Nipa tun ṣepọ awọn iwari igbẹkẹle kekere dipo sisọ wọn silẹ
   • Nipa yiyọ oluwari
   • Nipa lilo fireemu kan nikan
7. What does HOTA aim to balance in multi-object tracking evaluation?
   • Image resolution and model size
   • Training speed and GPU memory
   • Detection accuracy and association accuracy
   • Object color and camera frame rate
8. How does SORT assign detections to predicted tracks?
   • It uses a language model to name each object
   • It matches only by color histograms
   • It trains a new detector for each frame
   • It solves a frame-level assignment problem using IoU-based costs and the Hungarian algorithm