Track to Reconstruct and Reconstruct to Track

Luiten, Jonathon; Fischer, Tobias; Leibe, Bastian

doi:10.1109/lra.2020.2969183

Cited by 115 publications

(67 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This became the CLEAR MOT metrics (Bernardin and Stiefelhagen 2008) which positions the MOTA metric as the main metric for tracking evaluation alongside other metrics such as MOTP. MOTA was adopted for evaluation in the PETS workshop series (Ellis and Ferryman 2010) and remains, to this day, the most commonly used metric for evaluating MOT algorithms, although it has often been highly criticised (Shitrit et al 2011;Bento and Zhu 2016;Leichter and Krupka 2013;Leal-Taixé et al 2017;Milan et al 2013;Ristani et al 2016;Dave et al 2020;Luiten et al 2020;Maksai and Fua 2019;Wang et al 2019;Maksai et al 2017;Yu et al 2016;Dendorfer et al 2020;Luo et al 2014) for its bias toward overemphasizing detection over association (see Fig. 1), as well as a number of other issues (see Sect.…”

Section: Related Workmentioning

confidence: 99%

HOTA: A Higher Order Metric for Evaluating Multi-object Tracking

Luiten

Ošep²,

Dendorfer³

et al. 2020

Int J Comput Vis

Self Cite

641

234

View full text Add to dashboard Cite

Multi-object tracking (MOT) has been notoriously difficult to evaluate. Previous metrics overemphasize the importance of either detection or association. To address this, we present a novel MOT evaluation metric, higher order tracking accuracy (HOTA), which explicitly balances the effect of performing accurate detection, association and localization into a single unified metric for comparing trackers. HOTA decomposes into a family of sub-metrics which are able to evaluate each of five basic error types separately, which enables clear analysis of tracking performance. We evaluate the effectiveness of HOTA on the MOTChallenge benchmark, and show that it is able to capture important aspects of MOT performance not previously taken into account by established metrics. Furthermore, we show HOTA scores better align with human visual evaluation of tracking performance.

show abstract

Section: Related Workmentioning

confidence: 99%

HOTA: A Higher Order Metric for Evaluating Multi-object Tracking

Luiten

Ošep²,

Dendorfer³

et al. 2020

Int J Comput Vis

Self Cite

641

234

View full text Add to dashboard Cite

show abstract

“…Another deep learning approach which targets a similar problem is MOTSFusion (Luiten et al, 2019). This method aims to identify and separate cars, both still and moving.…”

Section: Related Workmentioning

confidence: 99%

Visual Odometryin Principle and Practice

Persson

2022

Linköping Studies in Science and Technology. Dissertations

View full text Add to dashboard Cite

Syn är det främsta sinnet genom vilket vi vet var vi är, vad som finns i närheten och hur vi rör oss. Motsvarande datorseendeuppgift är den simultana kartläggningen av omgivningen och lokaliseringen av kameran. Detta går under många namn varav jag använder visuell odometri, vilket innebär att bilderna är sekventiella och betonar noggrannheten i positioneringen och realtidskraven. Detta område har sett avsevärda förbättringar under det senaste decenniet och visuell odometri används flitigt inom robotik för lokalisering, navigering och hinderdetektion.Avhandlingens huvudsakliga syfte är att studera och utvecka visuella odometrisystem och ger flera bidrag. Det första är ett högpresterande stereo visuellt odometrisystem, som använde geometriskt stödd spårning för att uppnå högsta rank på KITTIs odometri tävling.Det andra är den ledande trepunktsperspetivlösaren. Lösaren hittar orientering och position för en kamera givet projektionerna av tre kända 3D-punkter och är en central del av många visuella odometrisystem. Genom att omformulera det underliggande problemet så undvek vi det problematiska fjärdegradspolynomet andra lösare lider av. Detta resulterade i avsevärt högre beräkningsprestanda och numerisk noggrannhet.Det tredje är ett system som generaliserar stereo visuell odometri till den samtidiga skattningen av flera oberoende rörliga objekt. Det huvudsakliga bidraget är ett realtidssystem som tillåter identifiering av rörliga stela objekt och skatting av deras banor i realtid, med tillämpningar inom robotnavigering i dynamiska miljöer.Det fjärde är ett förbättrad ramverk för skattning av kontinuerliga positioneringstrajektorier av splinetyp. Ramverket förenklar integrering av dynamiska modeller, och används för att visa att visuella odometrisystem baserade på kontinuerliga positioneringstrajectorier är både praktiska och kan köras i realtid.Visuella odometripipelinen betraktas både från ett teoretiskt och ett praktiskt perspektiv, och de designade systemen har testats på benchmarks och sensorplattformar. Den här avhandlingen placerar det publicerade arbetet i ett sammanhang och lyfter fram viktiga insikter och praktiska observationer.

show abstract

“…In [102], the authors proposed an object tracking and 3D reconstruction method to perform 3D object motion estimation. Object tracking and 3D reconstruction are often performed together, with tracking used as input for the 3D reconstruction.…”

Section: Tracking For Vehicle Reidmentioning

confidence: 99%

Re-Identification in Urban Scenarios: A Review of Tools and Methods

2021

View full text Add to dashboard Cite

With the widespread use of surveillance image cameras and enhanced awareness of public security, objects, and persons Re-Identification (ReID), the task of recognizing objects in non-overlapping camera networks has attracted particular attention in computer vision and pattern recognition communities. Given an image or video of an object-of-interest (query), object identification aims to identify the object from images or video feed taken from different cameras. After many years of great effort, object ReID remains a notably challenging task. The main reason is that an object’s appearance may dramatically change across camera views due to significant variations in illumination, poses or viewpoints, or even cluttered backgrounds. With the advent of Deep Neural Networks (DNN), there have been many proposals for different network architectures achieving high-performance levels. With the aim of identifying the most promising methods for ReID for future robust implementations, a review study is presented, mainly focusing on the person and multi-object ReID and auxiliary methods for image enhancement. Such methods are crucial for robust object ReID, while highlighting limitations of the identified methods. This is a very active field, evidenced by the dates of the publications found. However, most works use data from very different datasets and genres, which presents an obstacle to wide generalized DNN model training and usage. Although the model’s performance has achieved satisfactory results on particular datasets, a particular trend was observed in the use of 3D Convolutional Neural Networks (CNN), attention mechanisms to capture object-relevant features, and generative adversarial training to overcome data limitations. However, there is still room for improvement, namely in using images from urban scenarios among anonymized images to comply with public privacy legislation. The main challenges that remain in the ReID field, and prospects for future research directions towards ReID in dense urban scenarios, are also discussed.

show abstract

Track to Reconstruct and Reconstruct to Track

Cited by 115 publications

References 46 publications

HOTA: A Higher Order Metric for Evaluating Multi-object Tracking

HOTA: A Higher Order Metric for Evaluating Multi-object Tracking

Visual Odometryin Principle and Practice

Re-Identification in Urban Scenarios: A Review of Tools and Methods

Contact Info

Product

Resources

About