vehicle trajectory prediction in crowded highway
play

Vehicle Trajectory Prediction in Crowded Highway Scenarios Using - PowerPoint PPT Presentation

May 21, 2023 •137 likes •1.28k views

IEEE Intelligent Transportation Systems Conference - ITSC 2020 Vehicle Trajectory Prediction in Crowded Highway Scenarios Using Bird Eye View Representations and CNNs R. Izquierdo, A. Quintanar I. Parra, D. Fernndez-Llorca and M. A. Sotelo

  1. Dataset 6 HighD Dataset 1 • Publicy available at highd-dataset.com HighD Challenge • Road Structure Samples 39M 1500K • Unique IDs Vehicles 110K 110K Hours 147 16 • Vehicle dimension Clips 60 60 • Position, Speed, and Acceleration at 25 Hz 1 The highD Dataset: A Drone Dataset of Naturalistic Vehicle Trajectories on German Highways for Validation of Highly Automated Driving Systems. Krajewski, Robert and Bock, Julian and Kloeker, Laurent and Eckstein, Lutz R. Izquierdo | ITSC 2020, Virtual Conference

  2. Dataset 6 HighD Dataset 1 • Publicy available at highd-dataset.com HighD Challenge • Road Structure Samples 39M 1500K • Unique IDs Vehicles 110K 110K Hours 147 16 • Vehicle dimension Clips 60 60 • Position, Speed, and Acceleration at 25 Hz 1 The highD Dataset: A Drone Dataset of Naturalistic Vehicle Trajectories on German Highways for Validation of Highly Automated Driving Systems. Krajewski, Robert and Bock, Julian and Kloeker, Laurent and Eckstein, Lutz R. Izquierdo | ITSC 2020, Virtual Conference

  3. Dataset 6 HighD Dataset 1 • Publicy available at highd-dataset.com HighD Challenge • Road Structure Samples 39M 1500K • Unique IDs Vehicles 110K 110K Hours 147 16 • Vehicle dimension Clips 60 60 • Position, Speed, and Acceleration at 25 Hz 1 The highD Dataset: A Drone Dataset of Naturalistic Vehicle Trajectories on German Highways for Validation of Highly Automated Driving Systems. Krajewski, Robert and Bock, Julian and Kloeker, Laurent and Eckstein, Lutz R. Izquierdo | ITSC 2020, Virtual Conference

  4. Dataset 7 Data Codification • Driving Area as an Image: 32x512 m. → 64x512 px. • Vehicles as N ( µ, σ 2 ) • Possible Overlap P ( x , y ) = max {N i } ∀ i • Only Predictable Vehicles R. Izquierdo | ITSC 2020, Virtual Conference

  5. Dataset 7 Data Codification • Driving Area as an Image: 32x512 m. → 64x512 px. • Vehicles as N ( µ, σ 2 ) • Possible Overlap P ( x , y ) = max {N i } ∀ i • Only Predictable Vehicles R. Izquierdo | ITSC 2020, Virtual Conference

  6. Dataset 7 Data Codification • Driving Area as an Image: 32x512 m. → 64x512 px. • Vehicles as N ( µ, σ 2 ) • Possible Overlap P ( x , y ) = max {N i } ∀ i • Only Predictable Vehicles R. Izquierdo | ITSC 2020, Virtual Conference

  7. Dataset 7 Data Codification • Driving Area as an Image: 32x512 m. → 64x512 px. • Vehicles as N ( µ, σ 2 ) • Possible Overlap P ( x , y ) = max {N i } ∀ i • Only Predictable Vehicles R. Izquierdo | ITSC 2020, Virtual Conference

  8. Dataset 7 Data Codification • Driving Area as an Image: 32x512 m. → 64x512 px. • Vehicles as N ( µ, σ 2 ) • Possible Overlap P ( x , y ) = max {N i } ∀ i • Only Predictable Vehicles R. Izquierdo | ITSC 2020, Virtual Conference

  9. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  10. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  11. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  12. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  13. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  14. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  15. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  16. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  17. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  18. Dataset 8 Data Decodification • Iterative Extraction of Vehicle Positions • Search the pixel with the higest probability • Compute the mass center • Reset the area • Vehicle association by Euclidean distance R. Izquierdo | ITSC 2020, Virtual Conference

  19. Outline 9 Motivation Dataset Network Architecture Results Conclusions & Future Work R. Izquierdo | ITSC 2020, Virtual Conference

  20. Network Architecture 10 U-net model Depth levels 4 5 6 7 • Number of filters. Receptive Field ± 76 ± 156 ± 316 ± 636 Input size 16 32 64 128 • Depth levels. Parameters 56k 116k 235k 472k R. Izquierdo | ITSC 2020, Virtual Conference

  21. Network Architecture 10 U-net model Depth levels 4 5 6 7 • Number of filters. Receptive Field ± 76 ± 156 ± 316 ± 636 Input size 16 32 64 128 • Depth levels. Parameters 56k 116k 235k 472k R. Izquierdo | ITSC 2020, Virtual Conference

  22. Network Architecture 10 U-net model Depth levels 4 5 6 7 • Number of filters. Receptive Field ± 76 ± 156 ± 316 ± 636 Input size 16 32 64 128 • Depth levels. Parameters 56k 116k 235k 472k R. Izquierdo | ITSC 2020, Virtual Conference

  23. Network Architecture 10 U-net model Depth levels 4 5 6 7 • Number of filters. Receptive Field ± 76 ± 156 ± 316 ± 636 Input size 16 32 64 128 • Depth levels. Parameters 56k 116k 235k 472k R. Izquierdo | ITSC 2020, Virtual Conference

  24. Network Architecture 10 U-net model Depth levels 4 5 6 7 • Number of filters. Receptive Field ± 76 ± 156 ± 316 ± 636 Input size 16 32 64 128 • Depth levels. Parameters 56k 116k 235k 472k R. Izquierdo | ITSC 2020, Virtual Conference

  25. Network Architecture 10 U-net model Depth levels 4 5 6 7 • Number of filters. Receptive Field ± 76 ± 156 ± 316 ± 636 Input size 16 32 64 128 • Depth levels. Parameters 56k 116k 235k 472k R. Izquierdo | ITSC 2020, Virtual Conference

  26. Network Architecture 10 U-net model Depth levels 4 5 6 7 • Number of filters. Receptive Field ± 76 ± 156 ± 316 ± 636 Input size 16 32 64 128 • Depth levels. Parameters 56k 116k 235k 472k R. Izquierdo | ITSC 2020, Virtual Conference

  27. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  28. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  29. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  30. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  31. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  32. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  33. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  34. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  35. Network Architecture 11 Problem Application • Data rate lowered to 5 Hz • Input & Output 15@64x512 BEV − 2 . 8 ≤ t ≤ 3 . 0 • Activation Layer: • Linear • Clipped Rect. Linear Unit • Hyperbolic tangent • Loss function SE R. Izquierdo | ITSC 2020, Virtual Conference

  36. Outline 12 Motivation Dataset Network Architecture Results Conclusions & Future Work R. Izquierdo | ITSC 2020, Virtual Conference

  37. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  38. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  39. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  40. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  41. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  42. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  43. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  44. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  45. Results 13 Training setup • Seq 1-20 28K samples t = 0 . 2 t = 3 . 0 • Mini-batch size 1 Model ε x / ε y ε x / ε y • Epoch 1 Const. acc. 0.02 / 0.00 0.23 / 0.17 d = 5, f = linear 0.52 / 0.17 2.36 / 0.54 • Learning rate 10 − 6 d = 6, f = linear 0.23 / 0.01 1.23 / 0.07 • Momentum 0.9 d = 5, f = tanh - / - - / - • Loss function SE d = 6, f = tanh - / - - / - d = 5, f = cRelu 0.74 / 0.38 2.51 / 0.94 Test Results d = 6, f = cRelu 0.46 / 0.22 2.06 / 0.62 • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  46. Results 14 Training setup • Seq 1-20 28K samples • Mini-batch size 1 • Epoch 1 • Learning rate 10 − 6 • Momentum 0.9 • Loss function SE Test Results • Seq 21-25 7K samples • Position MAE R. Izquierdo | ITSC 2020, Virtual Conference

  47. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  48. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  49. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  50. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  51. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  52. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  53. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  54. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  55. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  56. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  57. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  58. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  59. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  60. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  61. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  62. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  63. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  64. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  65. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  66. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  67. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  68. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  69. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  70. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  71. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  72. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  73. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

  74. Results 15 Results Example Input block example Output block example ( × = GT + = Pred ) R. Izquierdo | ITSC 2020, Virtual Conference

Recommend

Pe Pedestria ian n Tra Trajectory jectory Predi redicti ction on Ov Overv rview

Pe Pedestria ian n Tra Trajectory jectory Predi redicti ction on Ov Overv rview

Pe Pedestria ian n Tra Trajectory jectory Predi redicti ction on Ov Overv rview What is pedestrian trajectory prediction good for? What is the task of pedestrian trajectory prediction? What makes trajectory prediction

925 views • 67 slides
DART: A Machine Learning Approach to Trajectory Prediction and Demand Capacity Balancing

DART: A Machine Learning Approach to Trajectory Prediction and Demand Capacity Balancing

DART: A Machine Learning Approach to Trajectory Prediction and Demand Capacity Balancing SESAR Belgrade, Serbia November 28 30 2017 Pablo Costas DART DART Project DART: Data driven Aircraft Trajectory prediction Research

229 views • 20 slides
Trajectory correction algorithms for a 3D underwater vehicle using affine transformations

Trajectory correction algorithms for a 3D underwater vehicle using affine transformations

Trajectory correction algorithms for a 3D underwater vehicle using affine transformations Quang-Cuong Pham and Yoshihiko Nakamura Nakamura-Takano Laboratory Department of Mechano-Informatics University of Tokyo Trajectory deformation

754 views • 47 slides
Trajectory (Motion) estimation of Autonomously Guided vehicle using Visual Odometry By Ashish

Trajectory (Motion) estimation of Autonomously Guided vehicle using Visual Odometry By Ashish

A Presentation on Trajectory (Motion) estimation of Autonomously Guided vehicle using Visual Odometry By Ashish Kumar M.Tech 1 st Year EE, IIT Kanpur Guide: Prof. Amitabha Mukherjee Subject : Artificial Intelligence ( CS365A ) Session:

273 views • 12 slides
Improving autonomous orchard vehicle trajectory tracking performance via slippage compensation

Improving autonomous orchard vehicle trajectory tracking performance via slippage compensation

Improving autonomous orchard vehicle trajectory tracking performance via slippage compensation Dr. Gokhan BAYAR Mechanical Engineering Department of Bulent Ecevit University Zonguldak, Turkey This study was conducted under the Supervision of

446 views • 43 slides
Techniques and Challenges for Trajectory Prediction Slides credit: Layla Pournajaf o Navigational

Techniques and Challenges for Trajectory Prediction Slides credit: Layla Pournajaf o Navigational

Where Next? Data Mining Techniques and Challenges for Trajectory Prediction Slides credit: Layla Pournajaf o Navigational services. o Traffic management. o Location-based advertising. Source: A. Monreale, F. Pinelli, R. Trasarti, F. Giannotti.

378 views • 37 slides
Vehicle Velocity Prediction Using Artificial Neural Networks and Effect of Real-World Signals on

Vehicle Velocity Prediction Using Artificial Neural Networks and Effect of Real-World Signals on

21-23 20 Vehicle Velocity Prediction Using Artificial Neural Networks and Effect of Real-World Signals on Prediction Window Tushar D. Gaikwad , Zachary D. Asher, Aaron Rabinowitz, Farhang Motallebiaraghi, Thomas Bradley, Alvis Fong, Rick Meyer

1.19k views • 27 slides
Vehicle Velocity Prediction Using Artificial Neural Networks and Effect of Real-World Signals on

Vehicle Velocity Prediction Using Artificial Neural Networks and Effect of Real-World Signals on

Vehicle Velocity Prediction Using Artificial Neural Networks and Effect of Real-World Signals on Prediction Window by, Tushar D. Gaikwad Committee Dr. Zachary Asher, Chair Dr. Richard Meyer Dr. Alvis Fong Western Michigan University 1

906 views • 65 slides
Newell Highway Upgrade Project 28 km of Newell Highway Single carriageway, single lane in

Newell Highway Upgrade Project 28 km of Newell Highway Single carriageway, single lane in

Newell Highway Upgrade Project 28 km of Newell Highway Single carriageway, single lane in each direction Highest heavy vehicle traffic in NSW Black soil country Annual flood risk Regional Geology Scope of

509 views • 27 slides
Lagranto 2.0 Contents An new object - trajectory Tutorial trajectory case study

Lagranto 2.0 Contents An new object - trajectory Tutorial trajectory case study

Lagranto 2.0 Contents An new object - trajectory Tutorial trajectory case study Goodies Documentation Libraries Installation Outlook Its the trajectory , stupid! Philospohy: There is a object called trajectory

655 views • 34 slides
Update on Proposed OHRV Regulatory Amendments California Off-Highway Motor Vehicle Recreation

Update on Proposed OHRV Regulatory Amendments California Off-Highway Motor Vehicle Recreation

Update on Proposed OHRV Regulatory Amendments California Off-Highway Motor Vehicle Recreation Commission Palmdale, California March 1, 2019 Why Does CARB Develop Regulations? California has the worst air quality in the nation We need

209 views • 17 slides
Comparing predefined and learned trajectory partitioning with applications to pedestrian route

Comparing predefined and learned trajectory partitioning with applications to pedestrian route

Comparing predefined and learned trajectory partitioning with applications to pedestrian route prediction Mark Dimond 1 , Gavin Smith 2 2 , James Goulding 2 , Mike Jackson 1 , Xiaolin Meng 1 1 Nottingham Geospatial Institute, University of

116 views • 8 slides
Impact of Traffic Lights on Trajectory Forecasting of Human-driven Vehicles Near Signalized

Impact of Traffic Lights on Trajectory Forecasting of Human-driven Vehicles Near Signalized

12th Workshop on Planning, Perception, Navigation for Intelligent Vehicle @ IROS2020 Impact of Traffic Lights on Trajectory Forecasting of Human-driven Vehicles Near Signalized Intersections Geunseob (GS) Oh , Huei Peng University of Michigan

372 views • 21 slides
Utilization of Crash and Medical Data to Reduce Motor Vehicle Crash Severity Funded with NHTSA

Utilization of Crash and Medical Data to Reduce Motor Vehicle Crash Severity Funded with NHTSA

Utilization of Crash and Medical Data to Reduce Motor Vehicle Crash Severity Funded with NHTSA Section 405-C funds through the Executive Office of Public Safety Highway Safety Divisions Highway Safety Division and the Traffic Records

550 views • 18 slides
Moving Object Trajectory Mining Moving Object Trajectory Mining Trajectory decomposition

Moving Object Trajectory Mining Moving Object Trajectory Mining Trajectory decomposition

Moving Object Trajectory Mining Moving Object Trajectory Mining Trajectory decomposition algorithm Trajectory decomposition algorithm Somayeh Dodge Robert Weibel Representation Analysis and Representation, Analysis and Visualization of

263 views • 10 slides
Trajectory planning Trajectory planning 1 1 Basilio Bona 1 ROBOTICA 03CFIOR

Trajectory planning Trajectory planning 1 1 Basilio Bona 1 ROBOTICA 03CFIOR

Trajectory planning Trajectory planning 1 1 Basilio Bona 1 ROBOTICA 03CFIOR Introduction The robot planning problem can be decomposed into a structured class of interconnected activities, at different hierarchical levels, usually

894 views • 54 slides
Article 19: PAYG Capital - Highway Ford Explorer $41,000 to replace existing 2005 Ford Explorer

Article 19: PAYG Capital - Highway Ford Explorer $41,000 to replace existing 2005 Ford Explorer

Article 19: PAYG Capital - Highway Ford Explorer $41,000 to replace existing 2005 Ford Explorer Frame is severely corroded and becoming structurally compromised D Vehicle failed inspection this year HOPKINTON Used 24/7 by Highway

69 views • 3 slides
Trajectory Code Validation j y Slides 04/12/08 04/12/08 AAE 450 Spring 2008 Trajectory

Trajectory Code Validation j y Slides 04/12/08 04/12/08 AAE 450 Spring 2008 Trajectory

Trajectory Code Validation j y Slides 04/12/08 04/12/08 AAE 450 Spring 2008 Trajectory Simulation Drag Loss Results: Results: Compare to Shuttle (2029633 kg GLOM): 107 m/s Titan IV/Centaur (886420 kg GLOM): 156 m/s (from

514 views • 22 slides
Crowded Field Photometry and Difference Imaging Przemek Wozniak Los Alamos National Laboratory

Crowded Field Photometry and Difference Imaging Przemek Wozniak Los Alamos National Laboratory

Crowded Field Photometry and Difference Imaging Przemek Wozniak Los Alamos National Laboratory Pasadena, Jul 2011 Sagan Exoplanet Workshop Przemek Wozniak Outline Motivation: Why crowded fields?

974 views • 41 slides
Highway Infrastructure Challenges Highway Infrastructure Challenges Presentation Ottawa,

Highway Infrastructure Challenges Highway Infrastructure Challenges Presentation Ottawa,

Highway Infrastructure Challenges Highway Infrastructure Challenges Presentation Ottawa, September 19, 2005 Canadian Trucking Alliance The National Highway System The National Highway System NHS comprises 25,000 kilometres of highway

305 views • 9 slides
PHYSICS OF BIOLOGICAL SYSTEMS ph549 LECTURE 14 Life in crowded environments Ideal vs.

PHYSICS OF BIOLOGICAL SYSTEMS ph549 LECTURE 14 Life in crowded environments Ideal vs.

PHYSICS OF BIOLOGICAL SYSTEMS ph549 LECTURE 14 Life in crowded environments Ideal vs. Real cells Rob Philips, Physical Biology of the Cell Protein Census in E. Coli Census in E. Coli Census in E. Coli The crowded Cytoskeleton Rob

594 views • 26 slides
Audiology Australias Events Portfolio Making room in a crowded space Dr Tony Coles Chief

Audiology Australias Events Portfolio Making room in a crowded space Dr Tony Coles Chief

Audiology Australias Events Portfolio Making room in a crowded space Dr Tony Coles Chief Executive Officer Overview Reliance on the National Conference A crowded space Mitigating risk Diversifying events Refreshing the National

231 views • 12 slides
V2X Technology Date: 11 th May 2013 What is V2X? V2X V2X - Vehicle to Vehicle or Vehicle to

V2X Technology Date: 11 th May 2013 What is V2X? V2X V2X - Vehicle to Vehicle or Vehicle to

V2X Technology Date: 11 th May 2013 What is V2X? V2X V2X - Vehicle to Vehicle or Vehicle to Infrastructure communication using wireless technology V2I V2V Vehicle to Vehicle to Vehicle Infrastructure Page 2 CONTENTS 1 Why V2X? 5

660 views • 16 slides
Non-Expert Development M A T T B U N T I N G , Y E G E T A Z E L E K E , K E N N O N M C K E

Non-Expert Development M A T T B U N T I N G , Y E G E T A Z E L E K E , K E N N O N M C K E

A Safe Autonomous Vehicle Trajectory Domain Specific Modeling Language For Non-Expert Development M A T T B U N T I N G , Y E G E T A Z E L E K E , K E N N O N M C K E E V E R , J O N A T H A N S P R I N K L E U N I V E R S I T Y O F A

465 views • 12 slides

More recommend