Correction to “Vid2Param: Modelling of Dynamics Parameters From Video”

Asenov, Martin; Burke, Michael G.; Angelov, Daniel; Davchev, Todor; Subr, Kartic; Ramamoorthy, Subramanian

doi:10.1109/lra.2020.2973022

Cited by 3 publications

(5 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Berdasarkan persamaan (7), nilai koefisien restitusi dapat diperoleh jika diketahui nilai dari perubahan ketinggian pantulan bouncing ball . Selain melalui persamaan (7), nilai dari koefisien restitusi secara teori juga dapat diperoleh melalui persamaan berikut,…”

Section: Hasil Dan Pembahasanunclassified

Analisis Parameter Dinamika Fenomena Bouncing Ball Berbasis Aplikasi Phyphox

Pebralia

2022

NMJ

View full text Add to dashboard Cite

Peristiwa bola yang memantul dapat digunakan sebagai model untuk mendeskripsikan banyak sekali aspek yang ada dalam mekanika. Pada penelitian ini dirancang fenomena bouncing ball dimana bola dijatuhkan dari ketinggian tertentu tanpa kecepatan awal dan menumbuk permukaan lantai yang keras. Penelitian ini menggunakan metode penelitian eksperimen menggunakan aplikasi Phyphox sebagai alat bantu untuk menentukan parameter-parameter yang ada pada fenomena bouncing ball . Adapun parameter yang dianalisis pada penelitian ini yaitu berupa kebergantungan energi, ketinggian, dan waktu. Hasil dari penelitian ini menunjukkan bahwa kecenderungan penurunan energi bouncing ball terhadap jumlah pantulan yang terjadi mengikuti grafik eksponensial. Pada eksperimen tahap I dengan bola kelereng berdiameter 15,2 mm diperoleh nilai penurunan energi tiap pantulan mengikuti persamaan eksponensial y=113,77 exp(-0,151x), dengan nilai . Pada eksperimen tahap II dengan bola kelereng berdiameter 15,2 mm diperoleh nilai penurunan energi tiap pantulan mengikuti persamaan eksponensial y=123,4 exp(-0,204x), dengan nilai.. Pada eksperimen tahap III dengan bola kelereng berdiameter 27,4 mm diperoleh nilai penurunan energi tiap pantulan mengikuti persamaan eksponensial y=115,01 exp(-0,176x), dengan nilai .

show abstract

Section: Hasil Dan Pembahasanunclassified

Analisis Parameter Dinamika Fenomena Bouncing Ball Berbasis Aplikasi Phyphox

Pebralia

2022

NMJ

View full text Add to dashboard Cite

show abstract

“…Recent work on physics-based deep learning injects structure in the latent space of the dynamics using Lagrangian and Hamiltonian operators (Greydanus et al, 2019;Chen et al, 2020;Toth et al, 2020;Sanchez-Gonzalez et al, 2019;Cranmer et al, 2020b;Zhong et al, 2020), by explicitly conserving physical quantities, or with ground truth supervision (Asenov et al, 2019;Wu et al, 2016;Xu et al, 2019b).…”

Section: Related Workmentioning

confidence: 99%

gradSim: Differentiable simulation for system identification and visuomotor control

Jatavallabhula,

Macklin,

Golemo

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Here, a bouncing ball scene with 2 moving distractors is shown, where the bouncing ball is correctly discovered as the object that corresponds to the highest entropy motion that fits the equation of motion f . ical parameter estimation [6,9], simulator learning [42,47], simulation alignment for robot interaction [2], trajectory generation [27] and compositionality [1,35].…”

Section: Related Workmentioning

confidence: 99%

“…CEM [45]; BO [39]). Estimation of the physical parameters over the full sequence (possibly hundreds of timesteps) is prone to local minima, as the dependency on the parameters can be highly non-linear 2 . This is further affected by the use of a non-optimized camera pose at the first iteration.…”

Section: Physical Parameter and Camera Pose Estimationmentioning

confidence: 99%

“…Proposal In an unlabeled video, ground-truth 2D keypoints are not available, but keypoint trajectories are required to maximize the likelihood in (2). Joint estimation of physical parameters with a neural network-based keypoint detector would be hard to optimize due to the difficulty of backpropagating through physics rollouts and camera projection into a CNN in a stable manner ( [24]).…”

Section: Trajectory Proposal and Selectionmentioning

confidence: 99%

See 1 more Smart Citation

Vision-based system identification and 3D keypoint discovery using dynamics constraints

Jaques,

Asenov,

Burke

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

This paper introduces V-SysId, a novel method that enables simultaneous keypoint discovery, 3D system identification, and extrinsic camera calibration from an unlabeled video taken from a static camera, using only the family of equations of motion of the object of interest as weak supervision. V-SysId takes keypoint trajectory proposals and alternates between maximum likelihood parameter estimation and extrinsic camera calibration, before applying a suitable selection criterion to identify the track of interest. This is then used to train a keypoint tracking model using supervised learning. Results on a range of settings (robotics, physics, physiology) highlight the utility of this approach.

show abstract

Correction to “Vid2Param: Modelling of Dynamics Parameters From Video”

Cited by 3 publications

References 1 publication

Analisis Parameter Dinamika Fenomena Bouncing Ball Berbasis Aplikasi Phyphox

Analisis Parameter Dinamika Fenomena Bouncing Ball Berbasis Aplikasi Phyphox

gradSim: Differentiable simulation for system identification and visuomotor control

Vision-based system identification and 3D keypoint discovery using dynamics constraints

Contact Info

Product

Resources

About