J. Bruske and G. Sommer. Intrinsic dimensionality extimation with optimally topology preserving maps. IEEE Trans. Pattern Analysis and Machine Intel- ligence, 20:572-575, 1998. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Gabor Wavelet Networks for Object Representation - Krueger (2002) (5 citations) (Correct)
....describe the flesh and hair color of the tracked person. The average error is claimed to be 6.8 (tilt) 5.7 (slant) and 2.9 (roll) but there was no investigation of stability. An appearance based approach similar to the one of [Schiele and Waibel, 1995] is investi gated in [Abraham Mumm, 1998; Bruske et al. 1998]. The approach allows computation of slant and tilt. The head is again tracked as a color blob. A square at the detected blob position in the image defines a region of interest (RAT) Within the RaT, complex 2 D Gabor filters (see eq. 2.17) are homogeneously distributed. Different filtering ....
....filters were applied and the range [0, r) of possible orientations was equidistantly sampled. The energies of the complex filter responses were fed into an ANN. A subspace variant of 108 the Local Linear Map (LLM) Ritter et al. 1991] is used as an ANN for learning the input output mapping [Bruske and Sommer, 1998]. The results were very promising; the reported mean errors were between 0.64 (4 x 4 filters, 4 orientations ( 0, and 7r) and 0.58 (8 x 8 filters, 4 orientations) see also Table 6.1) The errors were computed as follows: Let lb be the estimated slant, 0 the estimated tilt and let p and y be ....
[Article contains additional citation context not shown here]
J. Bruske and G. Sommer. Intrinsic dimensionality extimation with optimally topology preserving maps. IEEE Trans. Pattern Analysis and Machine Intel- ligence, 20:572-575, 1998.
Gabor Wavelet Networks for Efficient Head Pose Estimation - Krueger, Sommer Self-citation (Sommer) (Correct)
....3. 1 Experimental Setup In order to be comparable with the approach in [19] we used in our experiments the same neural network and the same number of training examples as described in [19] In [19] a subspace variant of the Local Linear Map (LLM) 22] was used for learning input output mappings [23]. There, the LLM rests on a locally linear (first order) approximation of the unknown function f : R and computes its output as (winner take all variant) y(x) A bmu (x c bmu ) o bmu . Here, o bmu is an output vector attached to the best matching unit (zero order approximation) and A ....
J. Bruske, G. Sommer, Intrinsic dimensionality extimation with optimally topology preserving maps, IEEE Trans. Pattern Analysis and Machine Intelligence 20 (1998) 572--575.
Efficient Head Pose Estimation with Gabor Wavelet Networks - Krüger, Bruns, Sommer (2000) (2 citations) Self-citation (Sommer) (Correct)
....wavelets (see g. 6) In order to be comparable with the approach in [1] we used in our experiments exactly the same neural network and the same number of training examples as described in [1] A subspace variant of the Local Linear Map (LLM) 16] was used for learning input output mappings [2]. The LLM rests on a locally linear ( rst order) approximation of the unknown function f : R n 7 R k and computes its output as (winner take all variant) y(x) Abmu (x c bmu ) obmu . Here, obmu 2 R k is an output vector attached to the best matching unit (zero order approximation) and ....
J. Bruske and G. Sommer. Intrinsic dimensionality extimation with optimally topology preserving maps. IEEE Trans. Pattern Analysis and Machine Intelligence, 20:572-575, 1998.
Efficient Head Pose Estimation with Gabor Wavelet Networks - Krüger, Bruns, Sommer (2000) (2 citations) Self-citation (Sommer) (Correct)
....wavelets (see fig. 6) In order to be comparable with the approach in [1] we used in our experiments exactly the same neural network and the same number of training examples as described in [1] A subspace variant of the Local Linear Map (LLM) 16] was used for learning input output mappings [2]. The LLM rests on a locally linear (first order) approximation of the unknown function f : R n 7 R k and computes its output as (winner take all variant) y(x) Abmu (x c bmu ) obmu . Here, o bmu 2 R k is an output vector attached to the best matching unit (zero order approximation) and ....
J. Bruske and G. Sommer. Intrinsic dimensionality extimation with optimally topology preserving maps. IEEE Trans. Pattern Analysis and Machine Intelligence, 20(5):572--575, 1998.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC