-
公开(公告)号：US20240330673A1
公开(公告)日：2024-10-03
申请号：US18184542
申请日：2023-03-15
Applicant: NavInfo Europe B.V.
Inventor： Kishaan Jeeveswaran , Prashant Shivaram Bhat , Elahe Arani , Bahram Zonooz
CPC classification number: G06N3/08 , G06V10/82 , G06V20/582
Abstract: A computer-implemented method for training a continual learning artificial neural network model, for sequential tasks, comprising an encoder and two classifiers. The method involves training the model on a plurality sequential tasks, with visual data being received from a vehicle mounted camera becoming increasingly available over time, wherein during each task, the model is presented with task-specific samples of the data and corresponding labels are drawn from a distribution. The method also includes adjusting the model on one task at a time, performing inference on all tasks previously encountered by said model, and using discrepancy loss between the two classifiers to regulate encoder and classifier weights adjustment so that the model adapts a representation cluster of samples from new tasks according to the clusters of previously learned tasks such that the model is suitable for a domain incremental learning scenario where tasks have shifting input distributions while the labels and/or classes remain the same.
-
32.发明授权公开(公告)号：US12062188B2
公开(公告)日：2024-08-13
申请号：US17691397
申请日：2022-03-10
Applicant: NavInfo Europe B.V.
CPC classification number: G06T7/20 , G06N3/088 , G06T2207/20081 , G06T2207/20084
Abstract: A computer implemented network for executing a self-supervised scene change detection method in which image pairs (T0, T1) from different time instances are subjected to random photometric transformations to obtain two pairs of augmented images (T0→T0′, T0″; T1→T1′, T1″), which augmented images are passed into an encoder (fθ) and a projection head (gϕ) to provide corresponding feature representations. Absolute feature differencing is applied over the outputs of the projection head (gϕ) to obtain difference representations (d1, d2) of changed features between the pair of images, and a self-supervised objective function (LSSL) is applied on the difference representations d1 and d2 to maximize a cross-correlation of the changed features, wherein d1 and d2 are defined as
d
1
=
❘
"\[LeftBracketingBar]"
g
(
f
(
T
0
′
)
)
-
g
(
f
(
T
0
′
)
)
❘
"\[RightBracketingBar]"
d
2
=
❘
"\[LeftBracketingBar]"
g
(
f
(
T
0
″
)
)
-
g
(
f
(
T
1
″
)
)
❘
"\[RightBracketingBar]"
(
1
)
Furthermore, an invariant prediction and change consistency loss is applied in the D-SSCD Network to reduce the effects of differences in the lighting conditions or camera viewpoints by enhancing the image alignment between the temporal images in the decision and feature space.-
公开(公告)号：US12061094B2
公开(公告)日：2024-08-13
申请号：US17674596
申请日：2022-02-17
Applicant: NavInfo Europe B.V.
Inventor： Haris Iqbal , Shruthi Gowda , Ahmed Badar , Terence Brouns , Arnav Varma , Elahe Arani , Bahram Zonooz
CPC classification number: G01C21/3815 , G01C21/32 , G06N3/08 , G06V20/582 , G06V20/588
Abstract: An AI based change detection system for executing a method to detect changes in geo-tagged videos to update HD maps, the method employing a neural network of modular components including a keyframe extraction module for processing two or more videos relating to separate traversals of an area of interest to which the HD map which is to be updated relates, a deep neural network module processing output of the keyframe extraction module, a change detection module processing output of the deep neural network module, and an auxiliary computations module which is designed to aid the change detection module.
-
公开(公告)号：US20240135169A1
公开(公告)日：2024-04-25
申请号：US18148257
申请日：2022-12-29
Applicant: NavInfo Europe B.V.
Inventor： Fahad Sarfraz , Elahe Arani , Bahram Zonooz
IPC: G06N3/08 , G06N3/0442 , G06N3/048
CPC classification number: G06N3/08 , G06N3/0442 , G06N3/048
Abstract: A computer-implemented method that encourages sparse coding in deep neural networks and mimics the interplay of multiple memory systems for maintaining a balance between stability and plasticity. To this end, the method includes a multi-memory experience replay mechanism that employs sparse coding. Activation sparsity is enforced along with a complementary dropout mechanism, which encourages the model to activate similar neurons for semantically similar inputs while reducing the overlap with activation patterns of semantically dissimilar inputs. The semantic dropout provides an efficient mechanism for balancing reusability and interference of features depending on the similarity of classes across tasks. Furthermore, the method includes the step of maintaining an additional long-term semantic memory that aggregates the information encoded in the synaptic weights of the working memory. An additional long-term semantic memory is maintained that aggregates the information encoded in the synaptic weights of the working memory.
-
公开(公告)号：US11948272B2
公开(公告)日：2024-04-02
申请号：US17402349
申请日：2021-08-13
Applicant: NavInfo Europe B.V.
Inventor： Hemang Chawla , Arnav Varma , Elahe Arani , Bahram Zonooz
IPC: G06T3/40 , G06F18/214 , G06N3/045 , G06N3/08 , G06T3/4038 , G06T3/4046 , G06T7/50 , G06V20/40
CPC classification number: G06T3/4046 , G06F18/2148 , G06N3/045 , G06N3/08 , G06T3/4038 , G06T7/50 , G06V20/41
Abstract: A computer-implemented method to improve scale consistency and/or scale awareness in a model of self-supervised depth and ego-motion prediction neural networks processing a video stream of monocular images, wherein complementary GPS coordinates synchronized with the images are used to calculate a GPS to scale loss to enforce the scale-consistency and/or -awareness on the monocular self-supervised ego-motion and depth estimation. A relative weight assigned to the GPS to scale loss exponentially increases as training progresses. The depth and ego-motion prediction neural networks are trained using an appearance-based photometric loss between real and synthesized target images, as well as a smoothness loss on the depth predictions.
-
Patent Agency Ranking
公开(公告)号：US20240037455A1
公开(公告)日：2024-02-01
申请号：US17894401
申请日：2022-08-24
Applicant: NavInfo Europe B.V.
Inventor： Naresh Kumar Gurulingan , Elahe Arani , Bahram Zonooz
CPC classification number: G06N20/10 , G06K9/6256 , G06K9/6215 , G06N3/063 , G06N5/022
Abstract: A computer-implemented method for multi-task structural learning in artificial neural network in which both the architecture and its parameters are learned simultaneously. The method utilizes two neural operators, namely, neuron creation and neuron removal, to aid in structural learning. The method creates excess neurons by starting from a disparate network for each task. Through the progress of training, corresponding task neurons in a layer pave the way for a specialized group neuron leading to a structural change. In the task learning phase of training, different neurons specialize in different tasks. In the interleaved structural learning phase, locally similar task neurons, before being removed, transfer their knowledge to a newly created group neuron. The training is completed with a final fine-tuning phase where only the multi-task loss is used.
-
37.发明公开公开(公告)号：US20230289977A1
公开(公告)日：2023-09-14
申请号：US17691397
申请日：2022-03-10
Applicant: NavInfo Europe B.V.
CPC classification number: G06T7/20 , G06N3/088 , G06T2207/20081 , G06T2207/20084
Abstract: A computer implemented network for executing a self-supervised scene change detection method in which image pairs (T0, T1) from different time instances are subjected to random photometric transformations to obtain two pairs of augmented images (T0 → T 0′, T 0‴ ; T1 → T 1′, T1″), which augmented images are passed into an encoder (fθ) and a projection head (gϕ) to provide corresponding feature representations. Absolute feature differencing is applied over the outputs of the projection head (gϕ) to obtain difference representations (d1, d2) of changed features between the pair of images, and a self-supervised objective function (LSSL) is applied on the difference representations d1 and d2 to maximize a cross-correlation of the changed features, wherein d1 and d2 are defined as
d
1
=
g
f
T
′
0
−
g
f
T
′
1
d
2
=
g
f
T
″
0
−
g
f
T
″
1
(1)
Furthermore, an invariant prediction and change consistency loss is applied in the D-SSCD Network to reduce the effects of differences in the lighting conditions or camera viewpoints by enhancing the image alignment between the temporal images in the decision and feature space.-
38.发明公开公开(公告)号：US20230258471A1
公开(公告)日：2023-08-17
申请号：US17674596
申请日：2022-02-17
Applicant: NavInfo Europe B.V.
Inventor： Haris Iqbal , Shruthi Gowda , Ahmed Badar , Terence Brouns , Arnav Varma , Elahe Arani , Bahram Zonooz
CPC classification number: G01C21/3815 , G06N3/08 , G01C21/32 , G06V20/588 , G06V20/582
Abstract: An AI based change detection system for executing a method to detect changes in geo-tagged videos to update HD maps, the method employing a neural network of modular components including a keyframe extraction module for processing two or more videos relating to separate traversals of an area of interest to which the HD map which is to be updated relates, a deep neural network module processing output of the keyframe extraction module, a change detection module processing output of the deep neural network module, and an auxiliary computations module which is designed to aid the change detection module.
-
公开(公告)号：US20220156882A1
公开(公告)日：2022-05-19
申请号：US17402349
申请日：2021-08-13
Applicant: NavInfo Europe B.V.
Inventor： Hemang Chawla , Arnav Varma , Elahe Arani , Bahram Zonooz
Abstract: A computer-implemented method to improve scale consistency and/or scale awareness in a model of self-supervised depth and ego-motion prediction neural networks processing a video stream of monocular images, wherein complementary GPS coordinates synchronized with the images are used to calculate a GPS to scale loss to enforce the scale-consistency and/or -awareness on the monocular self-supervised ego-motion and depth estimation. A relative weight assigned to the GPS to scale loss exponentially increases as training progresses. The depth and ego-motion prediction neural networks are trained using an appearance-based photometric loss between real and synthesized target images, as well as a smoothness loss on the depth predictions.
-
40.发明申请公开(公告)号：US20220092320A1
公开(公告)日：2022-03-24
申请号：US17482339
申请日：2021-09-22
Applicant: NavInfo Europe B.V.
Inventor： Terence Brouns , Elahe Arani , Bahram Zonooz
Abstract: A method and system for generating ground-truth annotations for object detection and classification for roadside objects in video data, wherein the method uses in combination an object detector to detect object instances of roadside objects in each frame of a video, a visual object tracker to detect and track the roadside object across the remaining video frames the roadside object appears in and clusters these detected object instances of the same roadside object into an object track, a trajectory analyzer to filter out object tracks that are unlikely from roadside objects, a classification model to classify each object instance in the object track into a predefined roadside object class, after which the object track as a whole is classified by seeking consensus among the individual object instance classifications in the object track, and classification consistency to determine whether the resulting roadside object class can be assigned automatically to the concerning object track as a ground-truth annotation or whether the ground-truth annotation should be manually verified by an operator. Accordingly, it is possible with the invention to convert model prediction labels in an automated way into ground-truth annotations, so as to create ground-truth annotations with a similar reliability as manual annotation and significantly reduce the amount of manual effort involved in creating reliable ground-truth annotations.
-
-
-
-
-
-
-
-
-
Search Results
42
records
(took 0.019 seconds)
